c1_math_0d_16s / trainer_state.json
neginr's picture
End of training
5109950 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.98632911392405,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 6.341896464804541,
"learning_rate": 3.2520325203252037e-07,
"loss": 0.8232,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 6.5313686335830266,
"learning_rate": 6.504065040650407e-07,
"loss": 0.8251,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 6.592492627336395,
"learning_rate": 9.75609756097561e-07,
"loss": 0.8439,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 6.07250791227797,
"learning_rate": 1.3008130081300815e-06,
"loss": 0.8142,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 5.921501476537587,
"learning_rate": 1.6260162601626018e-06,
"loss": 0.8163,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 5.796104673491665,
"learning_rate": 1.951219512195122e-06,
"loss": 0.82,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 4.335734306582742,
"learning_rate": 2.2764227642276426e-06,
"loss": 0.7768,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 4.22470622683716,
"learning_rate": 2.601626016260163e-06,
"loss": 0.7967,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.3533935215567463,
"learning_rate": 2.926829268292683e-06,
"loss": 0.7338,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.1078372830809835,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.7542,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.0173177093186023,
"learning_rate": 3.577235772357724e-06,
"loss": 0.7198,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.6548503282971723,
"learning_rate": 3.902439024390244e-06,
"loss": 0.702,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.276061996317132,
"learning_rate": 4.227642276422765e-06,
"loss": 0.744,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.033953367305103,
"learning_rate": 4.552845528455285e-06,
"loss": 0.7072,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 4.051143620135674,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.7169,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.14446723444474,
"learning_rate": 5.203252032520326e-06,
"loss": 0.6676,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.9624622919733796,
"learning_rate": 5.528455284552846e-06,
"loss": 0.6768,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.114172453223145,
"learning_rate": 5.853658536585366e-06,
"loss": 0.6446,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.493819216487882,
"learning_rate": 6.178861788617887e-06,
"loss": 0.6523,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.2661464833384606,
"learning_rate": 6.504065040650407e-06,
"loss": 0.6307,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.215086380299464,
"learning_rate": 6.829268292682928e-06,
"loss": 0.6152,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.3489315235368478,
"learning_rate": 7.154471544715448e-06,
"loss": 0.6102,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.0168098128003165,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.6174,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 0.7621930378831896,
"learning_rate": 7.804878048780489e-06,
"loss": 0.5836,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.8554007380913727,
"learning_rate": 8.130081300813009e-06,
"loss": 0.5603,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 0.8376767403709742,
"learning_rate": 8.45528455284553e-06,
"loss": 0.5767,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 0.6572490917912931,
"learning_rate": 8.78048780487805e-06,
"loss": 0.5715,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.8033170396222942,
"learning_rate": 9.10569105691057e-06,
"loss": 0.5634,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.7630803936669862,
"learning_rate": 9.43089430894309e-06,
"loss": 0.555,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.5796473989546078,
"learning_rate": 9.756097560975611e-06,
"loss": 0.5581,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.6837985203608627,
"learning_rate": 1.008130081300813e-05,
"loss": 0.5597,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.7906703322153873,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.5392,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.5329733577691343,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.5313,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.48718568532298223,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.5367,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.5977760978805214,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.5318,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.5174364521173804,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.531,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.48434892063454066,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.5346,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.5153799589717729,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.5249,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.3887592976909794,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.536,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.4435756068027413,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.5524,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.5740018765311709,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.5158,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.35902254257717026,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.514,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.45534959465877906,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.5024,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.42804371240070116,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.5126,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.3886234010240789,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.5165,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.450449635957086,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.5295,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.352296964273662,
"learning_rate": 1.528455284552846e-05,
"loss": 0.501,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.4659825712524389,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.5024,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.4688435593146043,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.5049,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.43422746514839267,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.5007,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.42349030647941427,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.4986,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.5139295079747652,
"learning_rate": 1.691056910569106e-05,
"loss": 0.5062,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.38091910850636285,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.4911,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.4944948009452858,
"learning_rate": 1.75609756097561e-05,
"loss": 0.5063,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.4582870951552393,
"learning_rate": 1.788617886178862e-05,
"loss": 0.5067,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.5692175638450153,
"learning_rate": 1.821138211382114e-05,
"loss": 0.4889,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.6084250125198883,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.4972,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.5123743829158213,
"learning_rate": 1.886178861788618e-05,
"loss": 0.4942,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.4019817856158885,
"learning_rate": 1.91869918699187e-05,
"loss": 0.4825,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.4805707446562215,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.4938,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.6180755493696108,
"learning_rate": 1.983739837398374e-05,
"loss": 0.5029,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.5576769070959305,
"learning_rate": 2.016260162601626e-05,
"loss": 0.4982,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.6528014700214161,
"learning_rate": 2.048780487804878e-05,
"loss": 0.4926,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.5324502326584601,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.5017,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.4721424144245652,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.4779,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.4608212029408293,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.4793,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.596340008380053,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.48,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.4602267129461295,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.4676,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.5981273519289902,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.4818,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.5255212766723283,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.4857,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.6673471969906832,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.4734,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.811793657410071,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.472,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.6824709324996017,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.4777,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.9221016396163514,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.4844,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.8552131816714748,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.4733,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.7956535764945188,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.466,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.9169635209836284,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.4738,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.7445290079880412,
"learning_rate": 2.536585365853659e-05,
"loss": 0.4739,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.7187460421306735,
"learning_rate": 2.569105691056911e-05,
"loss": 0.4686,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.7045888608711485,
"learning_rate": 2.601626016260163e-05,
"loss": 0.468,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.7874446046872324,
"learning_rate": 2.634146341463415e-05,
"loss": 0.4582,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.9237620036286267,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.4781,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.9669761209526861,
"learning_rate": 2.699186991869919e-05,
"loss": 0.476,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.0762583776433226,
"learning_rate": 2.731707317073171e-05,
"loss": 0.4767,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.735319929661224,
"learning_rate": 2.764227642276423e-05,
"loss": 0.4798,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.7708177963825833,
"learning_rate": 2.796747967479675e-05,
"loss": 0.4799,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.8670804580870681,
"learning_rate": 2.829268292682927e-05,
"loss": 0.4579,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.6482868010320487,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.4779,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.6591163084463784,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.4699,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.8024852545559771,
"learning_rate": 2.926829268292683e-05,
"loss": 0.4693,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.9629283993655922,
"learning_rate": 2.959349593495935e-05,
"loss": 0.4706,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.781249469590032,
"learning_rate": 2.991869918699187e-05,
"loss": 0.4747,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.6498563548001521,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.4651,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.7234379190347696,
"learning_rate": 3.056910569105692e-05,
"loss": 0.468,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.746555010681836,
"learning_rate": 3.089430894308943e-05,
"loss": 0.476,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.6865713693669165,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.4646,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.7834626689774069,
"learning_rate": 3.154471544715447e-05,
"loss": 0.4497,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.8865428337597154,
"learning_rate": 3.186991869918699e-05,
"loss": 0.4618,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 1.0096753544674797,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.4793,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.9265617156122333,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.4743,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.8857800505955531,
"learning_rate": 3.284552845528456e-05,
"loss": 0.4658,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.7847207963805071,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.4652,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.8348903419951106,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.4461,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.1515867743886607,
"learning_rate": 3.382113821138212e-05,
"loss": 0.471,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.8000949148571338,
"learning_rate": 3.414634146341463e-05,
"loss": 0.4751,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.8916184292447703,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.4638,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 1.051656522526483,
"learning_rate": 3.479674796747968e-05,
"loss": 0.4595,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.900171529882653,
"learning_rate": 3.51219512195122e-05,
"loss": 0.4571,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.7007611431480073,
"learning_rate": 3.544715447154472e-05,
"loss": 0.4599,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 0.7142697151437903,
"learning_rate": 3.577235772357724e-05,
"loss": 0.4587,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.780283658235145,
"learning_rate": 3.609756097560976e-05,
"loss": 0.4702,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.7357479237805824,
"learning_rate": 3.642276422764228e-05,
"loss": 0.4579,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 0.7954282049752563,
"learning_rate": 3.67479674796748e-05,
"loss": 0.4721,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.897158032160596,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.4651,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.007817442255216,
"learning_rate": 3.739837398373984e-05,
"loss": 0.4629,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.7996862598863689,
"learning_rate": 3.772357723577236e-05,
"loss": 0.4565,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 1.1834376325927127,
"learning_rate": 3.804878048780488e-05,
"loss": 0.461,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.950127536727282,
"learning_rate": 3.83739837398374e-05,
"loss": 0.4612,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.8916267666141113,
"learning_rate": 3.869918699186992e-05,
"loss": 0.4382,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 0.8260141646478527,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.4599,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.8228360217753147,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.4629,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.9390602819900682,
"learning_rate": 3.967479674796748e-05,
"loss": 0.4547,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.276033771081315,
"learning_rate": 4e-05,
"loss": 0.4657,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.5128746955411677,
"learning_rate": 3.999991946137476e-05,
"loss": 0.461,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.2052198714117122,
"learning_rate": 3.999967784614766e-05,
"loss": 0.4658,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.9817593402765697,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.4523,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.677351444496518,
"learning_rate": 3.999871139496895e-05,
"loss": 0.4458,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.8934769581538586,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.4724,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.9227621950897361,
"learning_rate": 3.999710067759846e-05,
"loss": 0.4628,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.565306250987119,
"learning_rate": 3.999605373449617e-05,
"loss": 0.4489,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.7306061048204809,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.4587,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.7332723936275084,
"learning_rate": 3.999347672161713e-05,
"loss": 0.444,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.6096518586674055,
"learning_rate": 3.999194667259528e-05,
"loss": 0.4536,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.7817642836220052,
"learning_rate": 3.999025561118334e-05,
"loss": 0.447,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.6508690945502666,
"learning_rate": 3.998840355100086e-05,
"loss": 0.4495,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.8691407111831984,
"learning_rate": 3.998639050696409e-05,
"loss": 0.4532,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.9196586188787509,
"learning_rate": 3.998421649528582e-05,
"loss": 0.4451,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.5744347704028048,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.4453,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.6985106341937295,
"learning_rate": 3.997938564033779e-05,
"loss": 0.4562,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.7186672251149245,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.4507,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.5778658567615854,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.4534,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.728958186189627,
"learning_rate": 3.997093258045973e-05,
"loss": 0.4484,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.6031137901675452,
"learning_rate": 3.996779317598936e-05,
"loss": 0.4671,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.623512151558159,
"learning_rate": 3.996449295365782e-05,
"loss": 0.4587,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.5672377646680865,
"learning_rate": 3.996103194004467e-05,
"loss": 0.4535,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.6230789442993361,
"learning_rate": 3.995741016302441e-05,
"loss": 0.4344,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.5640736617142171,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.4386,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.5016631201408571,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.4386,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.7283169893952158,
"learning_rate": 3.994558054968643e-05,
"loss": 0.4549,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.6542905341492657,
"learning_rate": 3.994131602367481e-05,
"loss": 0.449,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.5365034013979192,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.4328,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.5397555503049993,
"learning_rate": 3.993230519343752e-05,
"loss": 0.4395,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.49252157632522675,
"learning_rate": 3.992755896178383e-05,
"loss": 0.4506,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.4787928931531895,
"learning_rate": 3.992265223630981e-05,
"loss": 0.4274,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.512765117163987,
"learning_rate": 3.991758505653355e-05,
"loss": 0.4356,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.49642108852229605,
"learning_rate": 3.991235746326543e-05,
"loss": 0.4364,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.5837988129634571,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.4525,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.5805911814365201,
"learning_rate": 3.990142120595444e-05,
"loss": 0.4436,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.4076478497458924,
"learning_rate": 3.98957126299907e-05,
"loss": 0.4481,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.5838862930269725,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.4464,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.64528425118164,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.4521,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.4718065712416608,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.4529,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.5497083374979447,
"learning_rate": 3.98712764319087e-05,
"loss": 0.4419,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.48084338990062403,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.4378,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.46477249207814664,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.4293,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.5036194770080723,
"learning_rate": 3.985126869010129e-05,
"loss": 0.4365,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.5511341478585707,
"learning_rate": 3.984427961114833e-05,
"loss": 0.4522,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.70638298057329,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.4417,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.7082566147968332,
"learning_rate": 3.982982204151901e-05,
"loss": 0.4396,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.5892529075966988,
"learning_rate": 3.982235366728193e-05,
"loss": 0.445,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.5465871081996826,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.441,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.5636601147318336,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.4528,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.5421780720207464,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.4305,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.602192199274199,
"learning_rate": 3.979088432599627e-05,
"loss": 0.4313,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.4983084349065501,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.4285,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.4570453637730257,
"learning_rate": 3.977419304032111e-05,
"loss": 0.4412,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.4427169223958816,
"learning_rate": 3.976560847560697e-05,
"loss": 0.4402,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.4910339318302597,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.4312,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.5342796949600792,
"learning_rate": 3.974796184811956e-05,
"loss": 0.4316,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.5363161335874573,
"learning_rate": 3.973889992746979e-05,
"loss": 0.4433,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.5203016335076751,
"learning_rate": 3.972967903243361e-05,
"loss": 0.4333,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.5208154119335421,
"learning_rate": 3.972029923727486e-05,
"loss": 0.4337,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.610739009481546,
"learning_rate": 3.971076061753709e-05,
"loss": 0.4384,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.5205799875480199,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.4339,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.433052113051579,
"learning_rate": 3.969120721289402e-05,
"loss": 0.4355,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.49113949114153077,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.4317,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.4112703035712962,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.444,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.39062116386566564,
"learning_rate": 3.966068788369422e-05,
"loss": 0.4389,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.4188247863688021,
"learning_rate": 3.965019797448622e-05,
"loss": 0.4464,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.511093647630541,
"learning_rate": 3.963954980528515e-05,
"loss": 0.4439,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.49451973154349543,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.4174,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.4897605248515253,
"learning_rate": 3.961777903121329e-05,
"loss": 0.4395,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.43080821471027886,
"learning_rate": 3.960665660168131e-05,
"loss": 0.4288,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.5165645950527501,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.4371,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.6279005651375925,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.4282,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.36332480401041733,
"learning_rate": 3.957234222185657e-05,
"loss": 0.4265,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.5204942204229859,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.4218,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.7227907225701917,
"learning_rate": 3.954867765033605e-05,
"loss": 0.4353,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.6190347821844258,
"learning_rate": 3.953660915306728e-05,
"loss": 0.4333,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.528986857679542,
"learning_rate": 3.952438331063419e-05,
"loss": 0.4312,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.5982212595585554,
"learning_rate": 3.951200022150205e-05,
"loss": 0.4253,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.6977920027684116,
"learning_rate": 3.949945998540253e-05,
"loss": 0.4259,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.6967321962313783,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.4271,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.3668719130296035,
"learning_rate": 3.947390847755559e-05,
"loss": 0.4271,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.7255451198880996,
"learning_rate": 3.946089741159648e-05,
"loss": 0.4401,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.5830583625398662,
"learning_rate": 3.944772961024501e-05,
"loss": 0.4426,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.4971164094790236,
"learning_rate": 3.943440517955285e-05,
"loss": 0.4241,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.6561827292279085,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.4211,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.4432764001888977,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.4212,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.6205815998938581,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.4302,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.45141286344399906,
"learning_rate": 3.937954332854371e-05,
"loss": 0.4399,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.4652805917475414,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.4257,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.45109261704996306,
"learning_rate": 3.935117547697387e-05,
"loss": 0.4339,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.4179082305652065,
"learning_rate": 3.933675771619675e-05,
"loss": 0.4258,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.4897773699580438,
"learning_rate": 3.932218421983131e-05,
"loss": 0.418,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.6464075205042981,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.443,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.5940455249257359,
"learning_rate": 3.929257049108054e-05,
"loss": 0.4394,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.5646433230841752,
"learning_rate": 3.927753049720011e-05,
"loss": 0.4292,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.6638205628689277,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.4349,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.5799985060518066,
"learning_rate": 3.92469848560784e-05,
"loss": 0.4207,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.4979205603679978,
"learning_rate": 3.923147945484751e-05,
"loss": 0.4201,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.4080940950598775,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.438,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.5880503805880651,
"learning_rate": 3.920000411543654e-05,
"loss": 0.428,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.5393673618433,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.438,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.510669554048368,
"learning_rate": 3.916791024049648e-05,
"loss": 0.4353,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.4538018599113616,
"learning_rate": 3.91516316745245e-05,
"loss": 0.4286,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.48541134139102504,
"learning_rate": 3.913519886394389e-05,
"loss": 0.43,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.45939958416204174,
"learning_rate": 3.911861194110225e-05,
"loss": 0.4375,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.5448424579996259,
"learning_rate": 3.910187103958837e-05,
"loss": 0.4216,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.5915637180686979,
"learning_rate": 3.908497629423117e-05,
"loss": 0.4188,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.4753526057237399,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.4192,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.4653661629803726,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.4225,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.3854819242739982,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.4216,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.43338242011756384,
"learning_rate": 3.901586161429081e-05,
"loss": 0.4265,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.559297824123203,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.4319,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.3992708428154191,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.4222,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.4565981291935944,
"learning_rate": 3.896241703466667e-05,
"loss": 0.4229,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.4787243116114177,
"learning_rate": 3.894429654085585e-05,
"loss": 0.4324,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.46517493701947843,
"learning_rate": 3.892602347228505e-05,
"loss": 0.416,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.4506279391999084,
"learning_rate": 3.890759797612307e-05,
"loss": 0.4291,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.5239767525353708,
"learning_rate": 3.888902020076632e-05,
"loss": 0.4266,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.40184395329298633,
"learning_rate": 3.887029029583764e-05,
"loss": 0.4433,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.4877449993390723,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.43,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.6469850109060377,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.4396,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.5053142600393522,
"learning_rate": 3.881318931821972e-05,
"loss": 0.4198,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.4063557177471068,
"learning_rate": 3.879385241571817e-05,
"loss": 0.418,
"step": 246
},
{
"epoch": 1.0015189873417722,
"grad_norm": 0.6670763303655625,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.4099,
"step": 247
},
{
"epoch": 1.0055696202531645,
"grad_norm": 0.5219535619100215,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.3903,
"step": 248
},
{
"epoch": 1.009620253164557,
"grad_norm": 0.5608219991959764,
"learning_rate": 3.873493415863256e-05,
"loss": 0.3808,
"step": 249
},
{
"epoch": 1.0136708860759494,
"grad_norm": 0.5637696608702653,
"learning_rate": 3.871499275032111e-05,
"loss": 0.3868,
"step": 250
},
{
"epoch": 1.0177215189873419,
"grad_norm": 0.758739213874688,
"learning_rate": 3.869490061403091e-05,
"loss": 0.3932,
"step": 251
},
{
"epoch": 1.0217721518987342,
"grad_norm": 0.7402895114155325,
"learning_rate": 3.867465791158124e-05,
"loss": 0.3911,
"step": 252
},
{
"epoch": 1.0258227848101267,
"grad_norm": 0.522844632965814,
"learning_rate": 3.865426480600407e-05,
"loss": 0.3826,
"step": 253
},
{
"epoch": 1.029873417721519,
"grad_norm": 0.5657721622155955,
"learning_rate": 3.863372146154264e-05,
"loss": 0.3836,
"step": 254
},
{
"epoch": 1.0339240506329115,
"grad_norm": 0.7284303448245918,
"learning_rate": 3.861302804365024e-05,
"loss": 0.3922,
"step": 255
},
{
"epoch": 1.0379746835443038,
"grad_norm": 0.6749276065057042,
"learning_rate": 3.85921847189888e-05,
"loss": 0.3953,
"step": 256
},
{
"epoch": 1.0420253164556963,
"grad_norm": 0.4728172720970596,
"learning_rate": 3.85711916554276e-05,
"loss": 0.3958,
"step": 257
},
{
"epoch": 1.0460759493670886,
"grad_norm": 0.6178239514198169,
"learning_rate": 3.85500490220419e-05,
"loss": 0.385,
"step": 258
},
{
"epoch": 1.050126582278481,
"grad_norm": 0.5858316085063227,
"learning_rate": 3.852875698911154e-05,
"loss": 0.3863,
"step": 259
},
{
"epoch": 1.0541772151898734,
"grad_norm": 0.4759366347832318,
"learning_rate": 3.850731572811963e-05,
"loss": 0.3889,
"step": 260
},
{
"epoch": 1.058227848101266,
"grad_norm": 0.43515284748234573,
"learning_rate": 3.848572541175116e-05,
"loss": 0.3906,
"step": 261
},
{
"epoch": 1.0622784810126582,
"grad_norm": 0.5379866292984266,
"learning_rate": 3.846398621389154e-05,
"loss": 0.393,
"step": 262
},
{
"epoch": 1.0663291139240507,
"grad_norm": 0.5727467828835223,
"learning_rate": 3.84420983096253e-05,
"loss": 0.3862,
"step": 263
},
{
"epoch": 1.070379746835443,
"grad_norm": 0.3915861950143884,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.3912,
"step": 264
},
{
"epoch": 1.0744303797468355,
"grad_norm": 0.41739987667891626,
"learning_rate": 3.839787708819787e-05,
"loss": 0.3965,
"step": 265
},
{
"epoch": 1.0784810126582278,
"grad_norm": 0.38227951920031317,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.3927,
"step": 266
},
{
"epoch": 1.0825316455696203,
"grad_norm": 0.3685809210020584,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.408,
"step": 267
},
{
"epoch": 1.0865822784810126,
"grad_norm": 0.4077992189938042,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.3867,
"step": 268
},
{
"epoch": 1.0906329113924051,
"grad_norm": 0.4020174365143566,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.3854,
"step": 269
},
{
"epoch": 1.0946835443037974,
"grad_norm": 0.36713279943040567,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.3864,
"step": 270
},
{
"epoch": 1.09873417721519,
"grad_norm": 0.4298163243433075,
"learning_rate": 3.826166304956594e-05,
"loss": 0.3889,
"step": 271
},
{
"epoch": 1.1027848101265822,
"grad_norm": 0.3791077845183942,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.3892,
"step": 272
},
{
"epoch": 1.1068354430379748,
"grad_norm": 0.34006625966761617,
"learning_rate": 3.821507978767479e-05,
"loss": 0.3842,
"step": 273
},
{
"epoch": 1.110886075949367,
"grad_norm": 0.40288735275498155,
"learning_rate": 3.819156801001693e-05,
"loss": 0.391,
"step": 274
},
{
"epoch": 1.1149367088607596,
"grad_norm": 0.3778423111283362,
"learning_rate": 3.816790971997121e-05,
"loss": 0.3673,
"step": 275
},
{
"epoch": 1.1189873417721519,
"grad_norm": 0.3906045522877888,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.3827,
"step": 276
},
{
"epoch": 1.1230379746835444,
"grad_norm": 0.35014849619477395,
"learning_rate": 3.81201543660571e-05,
"loss": 0.3862,
"step": 277
},
{
"epoch": 1.1270886075949367,
"grad_norm": 0.44245455503187525,
"learning_rate": 3.809605768680377e-05,
"loss": 0.3775,
"step": 278
},
{
"epoch": 1.1311392405063292,
"grad_norm": 0.4002112621047507,
"learning_rate": 3.807181526438958e-05,
"loss": 0.3987,
"step": 279
},
{
"epoch": 1.1351898734177215,
"grad_norm": 0.32247254709723555,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.3794,
"step": 280
},
{
"epoch": 1.139240506329114,
"grad_norm": 0.3641819455798482,
"learning_rate": 3.802289397223145e-05,
"loss": 0.3879,
"step": 281
},
{
"epoch": 1.1432911392405063,
"grad_norm": 0.34900682615055734,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.3924,
"step": 282
},
{
"epoch": 1.1473417721518988,
"grad_norm": 0.3209179684008272,
"learning_rate": 3.797339206560096e-05,
"loss": 0.3844,
"step": 283
},
{
"epoch": 1.151392405063291,
"grad_norm": 0.35468433314099324,
"learning_rate": 3.794842387948027e-05,
"loss": 0.381,
"step": 284
},
{
"epoch": 1.1554430379746836,
"grad_norm": 0.3185632664498542,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.3685,
"step": 285
},
{
"epoch": 1.159493670886076,
"grad_norm": 0.3664130722426212,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.3787,
"step": 286
},
{
"epoch": 1.1635443037974684,
"grad_norm": 0.43468048736416975,
"learning_rate": 3.787265280646825e-05,
"loss": 0.385,
"step": 287
},
{
"epoch": 1.1675949367088607,
"grad_norm": 0.4011551052192431,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.3986,
"step": 288
},
{
"epoch": 1.1716455696202532,
"grad_norm": 0.3125851588154434,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.3809,
"step": 289
},
{
"epoch": 1.1756962025316455,
"grad_norm": 0.3602141621630121,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.3868,
"step": 290
},
{
"epoch": 1.179746835443038,
"grad_norm": 0.37102524990677505,
"learning_rate": 3.776961046830306e-05,
"loss": 0.4017,
"step": 291
},
{
"epoch": 1.1837974683544303,
"grad_norm": 0.353145072880679,
"learning_rate": 3.774349157719215e-05,
"loss": 0.3957,
"step": 292
},
{
"epoch": 1.1878481012658229,
"grad_norm": 0.3928767504407124,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.3898,
"step": 293
},
{
"epoch": 1.1918987341772151,
"grad_norm": 0.3545909346245163,
"learning_rate": 3.769082529555359e-05,
"loss": 0.3853,
"step": 294
},
{
"epoch": 1.1959493670886077,
"grad_norm": 0.36477200092725864,
"learning_rate": 3.766427832919294e-05,
"loss": 0.3845,
"step": 295
},
{
"epoch": 1.2,
"grad_norm": 0.37836497881065845,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.3885,
"step": 296
},
{
"epoch": 1.2040506329113925,
"grad_norm": 0.3596042822426635,
"learning_rate": 3.761075781441526e-05,
"loss": 0.3932,
"step": 297
},
{
"epoch": 1.2081012658227848,
"grad_norm": 0.3455402908245076,
"learning_rate": 3.75837846970451e-05,
"loss": 0.3941,
"step": 298
},
{
"epoch": 1.2121518987341773,
"grad_norm": 0.35658530616496864,
"learning_rate": 3.755666996229032e-05,
"loss": 0.3708,
"step": 299
},
{
"epoch": 1.2162025316455696,
"grad_norm": 0.4447056367690691,
"learning_rate": 3.752941382852927e-05,
"loss": 0.3827,
"step": 300
},
{
"epoch": 1.220253164556962,
"grad_norm": 0.46694737892782073,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.3856,
"step": 301
},
{
"epoch": 1.2243037974683544,
"grad_norm": 0.37325891047542203,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.3871,
"step": 302
},
{
"epoch": 1.228354430379747,
"grad_norm": 0.43446357064177793,
"learning_rate": 3.744679923406351e-05,
"loss": 0.3798,
"step": 303
},
{
"epoch": 1.2324050632911392,
"grad_norm": 0.4144350535211654,
"learning_rate": 3.741897971081043e-05,
"loss": 0.3976,
"step": 304
},
{
"epoch": 1.2364556962025317,
"grad_norm": 0.5033983976341748,
"learning_rate": 3.739101989748946e-05,
"loss": 0.3792,
"step": 305
},
{
"epoch": 1.240506329113924,
"grad_norm": 0.4409292512937945,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.3859,
"step": 306
},
{
"epoch": 1.2445569620253165,
"grad_norm": 0.4078054563418798,
"learning_rate": 3.73346803025098e-05,
"loss": 0.3876,
"step": 307
},
{
"epoch": 1.2486075949367088,
"grad_norm": 0.5726628314300202,
"learning_rate": 3.730630097460247e-05,
"loss": 0.3845,
"step": 308
},
{
"epoch": 1.2526582278481013,
"grad_norm": 0.40986200047897253,
"learning_rate": 3.727778226412628e-05,
"loss": 0.3889,
"step": 309
},
{
"epoch": 1.2567088607594936,
"grad_norm": 0.4150534281513503,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.3898,
"step": 310
},
{
"epoch": 1.2607594936708861,
"grad_norm": 0.40741193692267197,
"learning_rate": 3.722032761533114e-05,
"loss": 0.3891,
"step": 311
},
{
"epoch": 1.2648101265822784,
"grad_norm": 0.478679572791949,
"learning_rate": 3.719139213974403e-05,
"loss": 0.3765,
"step": 312
},
{
"epoch": 1.268860759493671,
"grad_norm": 0.3893561287928256,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.3912,
"step": 313
},
{
"epoch": 1.2729113924050632,
"grad_norm": 0.4293031819026889,
"learning_rate": 3.713310605140055e-05,
"loss": 0.384,
"step": 314
},
{
"epoch": 1.2769620253164558,
"grad_norm": 0.5746079111255373,
"learning_rate": 3.710375590807233e-05,
"loss": 0.4004,
"step": 315
},
{
"epoch": 1.281012658227848,
"grad_norm": 0.5110550411358974,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.3704,
"step": 316
},
{
"epoch": 1.2850632911392406,
"grad_norm": 0.35925283157656573,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.3749,
"step": 317
},
{
"epoch": 1.2891139240506329,
"grad_norm": 0.4349524322162539,
"learning_rate": 3.701487992136854e-05,
"loss": 0.3941,
"step": 318
},
{
"epoch": 1.2931645569620254,
"grad_norm": 0.46120690850782375,
"learning_rate": 3.69849802022222e-05,
"loss": 0.3859,
"step": 319
},
{
"epoch": 1.2972151898734177,
"grad_norm": 0.40213658014601467,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.3796,
"step": 320
},
{
"epoch": 1.3012658227848102,
"grad_norm": 0.4333565489628601,
"learning_rate": 3.692477062175289e-05,
"loss": 0.3849,
"step": 321
},
{
"epoch": 1.3053164556962025,
"grad_norm": 0.48325989849083656,
"learning_rate": 3.689446124534958e-05,
"loss": 0.3913,
"step": 322
},
{
"epoch": 1.309367088607595,
"grad_norm": 0.4564175029903682,
"learning_rate": 3.686401580327799e-05,
"loss": 0.3885,
"step": 323
},
{
"epoch": 1.3134177215189873,
"grad_norm": 0.4612241023143642,
"learning_rate": 3.683343454074149e-05,
"loss": 0.383,
"step": 324
},
{
"epoch": 1.3174683544303798,
"grad_norm": 0.4316264989500914,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.385,
"step": 325
},
{
"epoch": 1.321518987341772,
"grad_norm": 0.41730502150534116,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.3882,
"step": 326
},
{
"epoch": 1.3255696202531646,
"grad_norm": 0.46701853188364373,
"learning_rate": 3.674087829877297e-05,
"loss": 0.3852,
"step": 327
},
{
"epoch": 1.329620253164557,
"grad_norm": 0.4071533637265324,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.3831,
"step": 328
},
{
"epoch": 1.3336708860759494,
"grad_norm": 0.4275155341281297,
"learning_rate": 3.667849957966501e-05,
"loss": 0.3875,
"step": 329
},
{
"epoch": 1.3377215189873417,
"grad_norm": 0.39290503692832546,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.3785,
"step": 330
},
{
"epoch": 1.3417721518987342,
"grad_norm": 0.3971452241428047,
"learning_rate": 3.661558355626795e-05,
"loss": 0.4002,
"step": 331
},
{
"epoch": 1.3458227848101265,
"grad_norm": 0.529237671710245,
"learning_rate": 3.658392468818163e-05,
"loss": 0.3896,
"step": 332
},
{
"epoch": 1.349873417721519,
"grad_norm": 0.3385900810296459,
"learning_rate": 3.655213225544574e-05,
"loss": 0.3897,
"step": 333
},
{
"epoch": 1.3539240506329113,
"grad_norm": 0.3847536988596601,
"learning_rate": 3.652020651411218e-05,
"loss": 0.3872,
"step": 334
},
{
"epoch": 1.3579746835443038,
"grad_norm": 0.5080750546772402,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.3952,
"step": 335
},
{
"epoch": 1.3620253164556961,
"grad_norm": 0.3295359414894108,
"learning_rate": 3.645595613522574e-05,
"loss": 0.3767,
"step": 336
},
{
"epoch": 1.3660759493670886,
"grad_norm": 0.3912443666371073,
"learning_rate": 3.642363201513657e-05,
"loss": 0.3882,
"step": 337
},
{
"epoch": 1.370126582278481,
"grad_norm": 0.459935450978153,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.4066,
"step": 338
},
{
"epoch": 1.3741772151898735,
"grad_norm": 0.4841854463691737,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.3747,
"step": 339
},
{
"epoch": 1.3782278481012658,
"grad_norm": 0.41504970158254095,
"learning_rate": 3.632586705948318e-05,
"loss": 0.3742,
"step": 340
},
{
"epoch": 1.3822784810126583,
"grad_norm": 0.39966500729354626,
"learning_rate": 3.629301541734311e-05,
"loss": 0.3964,
"step": 341
},
{
"epoch": 1.3863291139240506,
"grad_norm": 0.4962014539984421,
"learning_rate": 3.626003255349676e-05,
"loss": 0.3848,
"step": 342
},
{
"epoch": 1.390379746835443,
"grad_norm": 0.4545545029216032,
"learning_rate": 3.622691873358357e-05,
"loss": 0.391,
"step": 343
},
{
"epoch": 1.3944303797468354,
"grad_norm": 0.34293200793640255,
"learning_rate": 3.61936742242977e-05,
"loss": 0.3926,
"step": 344
},
{
"epoch": 1.3984810126582279,
"grad_norm": 0.3880042788536112,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.3698,
"step": 345
},
{
"epoch": 1.4025316455696202,
"grad_norm": 0.458294865196501,
"learning_rate": 3.612679420964516e-05,
"loss": 0.4005,
"step": 346
},
{
"epoch": 1.4065822784810127,
"grad_norm": 0.3250273298203594,
"learning_rate": 3.609315924292092e-05,
"loss": 0.3845,
"step": 347
},
{
"epoch": 1.410632911392405,
"grad_norm": 0.3341994888885682,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.3927,
"step": 348
},
{
"epoch": 1.4146835443037975,
"grad_norm": 0.3214155737544187,
"learning_rate": 3.602550074513133e-05,
"loss": 0.388,
"step": 349
},
{
"epoch": 1.4187341772151898,
"grad_norm": 0.2974408531468748,
"learning_rate": 3.599147775897822e-05,
"loss": 0.3872,
"step": 350
},
{
"epoch": 1.4227848101265823,
"grad_norm": 0.39233747440826344,
"learning_rate": 3.595732597966167e-05,
"loss": 0.387,
"step": 351
},
{
"epoch": 1.4268354430379746,
"grad_norm": 0.3153010739466784,
"learning_rate": 3.592304568223542e-05,
"loss": 0.3881,
"step": 352
},
{
"epoch": 1.4308860759493671,
"grad_norm": 0.3621304820250998,
"learning_rate": 3.588863714278826e-05,
"loss": 0.3789,
"step": 353
},
{
"epoch": 1.4349367088607594,
"grad_norm": 0.39513336117237713,
"learning_rate": 3.585410063844186e-05,
"loss": 0.3808,
"step": 354
},
{
"epoch": 1.438987341772152,
"grad_norm": 0.41358331777651786,
"learning_rate": 3.581943644734846e-05,
"loss": 0.3917,
"step": 355
},
{
"epoch": 1.4430379746835442,
"grad_norm": 0.35929762564090645,
"learning_rate": 3.578464484868869e-05,
"loss": 0.3846,
"step": 356
},
{
"epoch": 1.4470886075949367,
"grad_norm": 0.3757377672633556,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.3875,
"step": 357
},
{
"epoch": 1.451139240506329,
"grad_norm": 0.4179276227781696,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.3843,
"step": 358
},
{
"epoch": 1.4551898734177215,
"grad_norm": 0.32796873502136226,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.3817,
"step": 359
},
{
"epoch": 1.459240506329114,
"grad_norm": 0.3294321990067231,
"learning_rate": 3.564420999786543e-05,
"loss": 0.3774,
"step": 360
},
{
"epoch": 1.4632911392405064,
"grad_norm": 0.37601646974923375,
"learning_rate": 3.560878558491842e-05,
"loss": 0.3903,
"step": 361
},
{
"epoch": 1.4673417721518986,
"grad_norm": 0.3097264221346817,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.3961,
"step": 362
},
{
"epoch": 1.4713924050632912,
"grad_norm": 0.312917389872051,
"learning_rate": 3.553755991230039e-05,
"loss": 0.3747,
"step": 363
},
{
"epoch": 1.4754430379746837,
"grad_norm": 0.3392810720233997,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.3854,
"step": 364
},
{
"epoch": 1.479493670886076,
"grad_norm": 0.42146159952968454,
"learning_rate": 3.546583369120419e-05,
"loss": 0.3755,
"step": 365
},
{
"epoch": 1.4835443037974683,
"grad_norm": 0.3557028111368005,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.3901,
"step": 366
},
{
"epoch": 1.4875949367088608,
"grad_norm": 0.3181209210357592,
"learning_rate": 3.539360923231766e-05,
"loss": 0.3884,
"step": 367
},
{
"epoch": 1.4916455696202533,
"grad_norm": 0.41350536486871536,
"learning_rate": 3.535731089018394e-05,
"loss": 0.3991,
"step": 368
},
{
"epoch": 1.4956962025316456,
"grad_norm": 0.3940955007612258,
"learning_rate": 3.532088886237956e-05,
"loss": 0.3801,
"step": 369
},
{
"epoch": 1.4997468354430379,
"grad_norm": 0.34028464161617367,
"learning_rate": 3.528434344224253e-05,
"loss": 0.3746,
"step": 370
},
{
"epoch": 1.5037974683544304,
"grad_norm": 0.3814866552810181,
"learning_rate": 3.524767492410464e-05,
"loss": 0.3756,
"step": 371
},
{
"epoch": 1.507848101265823,
"grad_norm": 0.38186712774813913,
"learning_rate": 3.521088360328908e-05,
"loss": 0.3937,
"step": 372
},
{
"epoch": 1.5118987341772152,
"grad_norm": 0.3185112635061116,
"learning_rate": 3.517396977610811e-05,
"loss": 0.3848,
"step": 373
},
{
"epoch": 1.5159493670886075,
"grad_norm": 0.3496781917827941,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.4022,
"step": 374
},
{
"epoch": 1.52,
"grad_norm": 0.41230794244929525,
"learning_rate": 3.509977579282971e-05,
"loss": 0.3827,
"step": 375
},
{
"epoch": 1.5240506329113925,
"grad_norm": 0.36184039425743675,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.3714,
"step": 376
},
{
"epoch": 1.5281012658227848,
"grad_norm": 0.345325051749899,
"learning_rate": 3.502509536445719e-05,
"loss": 0.3816,
"step": 377
},
{
"epoch": 1.5321518987341771,
"grad_norm": 0.35295911728445606,
"learning_rate": 3.498757348458147e-05,
"loss": 0.3741,
"step": 378
},
{
"epoch": 1.5362025316455696,
"grad_norm": 0.33136209253260407,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.3795,
"step": 379
},
{
"epoch": 1.5402531645569622,
"grad_norm": 0.34123752571172705,
"learning_rate": 3.491216790442899e-05,
"loss": 0.3726,
"step": 380
},
{
"epoch": 1.5443037974683544,
"grad_norm": 0.3766340722619941,
"learning_rate": 3.487428481145839e-05,
"loss": 0.3842,
"step": 381
},
{
"epoch": 1.5483544303797467,
"grad_norm": 0.34106773161005055,
"learning_rate": 3.483628192304278e-05,
"loss": 0.4019,
"step": 382
},
{
"epoch": 1.5524050632911393,
"grad_norm": 0.2775183322137869,
"learning_rate": 3.479815954525219e-05,
"loss": 0.3718,
"step": 383
},
{
"epoch": 1.5564556962025318,
"grad_norm": 0.36222179313147235,
"learning_rate": 3.475991798511899e-05,
"loss": 0.3914,
"step": 384
},
{
"epoch": 1.560506329113924,
"grad_norm": 0.32216458320882624,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.3828,
"step": 385
},
{
"epoch": 1.5645569620253164,
"grad_norm": 0.3268956328759394,
"learning_rate": 3.468307855075128e-05,
"loss": 0.3846,
"step": 386
},
{
"epoch": 1.5686075949367089,
"grad_norm": 0.3645092332274162,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.3795,
"step": 387
},
{
"epoch": 1.5726582278481014,
"grad_norm": 0.322553057143406,
"learning_rate": 3.460576609535163e-05,
"loss": 0.3774,
"step": 388
},
{
"epoch": 1.5767088607594937,
"grad_norm": 0.32215087756490307,
"learning_rate": 3.456693326250006e-05,
"loss": 0.3869,
"step": 389
},
{
"epoch": 1.580759493670886,
"grad_norm": 0.2988708469288339,
"learning_rate": 3.452798310957058e-05,
"loss": 0.3795,
"step": 390
},
{
"epoch": 1.5848101265822785,
"grad_norm": 0.3402631526483103,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.3786,
"step": 391
},
{
"epoch": 1.588860759493671,
"grad_norm": 0.3060490347445895,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.3748,
"step": 392
},
{
"epoch": 1.5929113924050633,
"grad_norm": 0.3236796323312587,
"learning_rate": 3.441043187201574e-05,
"loss": 0.3926,
"step": 393
},
{
"epoch": 1.5969620253164556,
"grad_norm": 0.3609103351482647,
"learning_rate": 3.437101558517728e-05,
"loss": 0.3892,
"step": 394
},
{
"epoch": 1.601012658227848,
"grad_norm": 0.3804288106279976,
"learning_rate": 3.433148355615496e-05,
"loss": 0.3874,
"step": 395
},
{
"epoch": 1.6050632911392406,
"grad_norm": 0.3990776439602732,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.3795,
"step": 396
},
{
"epoch": 1.609113924050633,
"grad_norm": 0.3108741442246338,
"learning_rate": 3.425207354603043e-05,
"loss": 0.3696,
"step": 397
},
{
"epoch": 1.6131645569620252,
"grad_norm": 0.4023502984494289,
"learning_rate": 3.421219620448553e-05,
"loss": 0.3802,
"step": 398
},
{
"epoch": 1.6172151898734177,
"grad_norm": 0.3473046854521834,
"learning_rate": 3.417220439986623e-05,
"loss": 0.3822,
"step": 399
},
{
"epoch": 1.6212658227848102,
"grad_norm": 0.31466254640093355,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.3774,
"step": 400
},
{
"epoch": 1.6253164556962025,
"grad_norm": 0.3591169745638006,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.3745,
"step": 401
},
{
"epoch": 1.6293670886075948,
"grad_norm": 0.3493127569664842,
"learning_rate": 3.405154543304065e-05,
"loss": 0.3776,
"step": 402
},
{
"epoch": 1.6334177215189873,
"grad_norm": 0.3619705811333053,
"learning_rate": 3.401109900618843e-05,
"loss": 0.3703,
"step": 403
},
{
"epoch": 1.6374683544303799,
"grad_norm": 0.3969500100350988,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.3755,
"step": 404
},
{
"epoch": 1.6415189873417722,
"grad_norm": 0.47561935207942,
"learning_rate": 3.392986794874714e-05,
"loss": 0.3779,
"step": 405
},
{
"epoch": 1.6455696202531644,
"grad_norm": 0.3391110302578449,
"learning_rate": 3.388908397238184e-05,
"loss": 0.3825,
"step": 406
},
{
"epoch": 1.649620253164557,
"grad_norm": 0.460091674281835,
"learning_rate": 3.384818813524362e-05,
"loss": 0.3835,
"step": 407
},
{
"epoch": 1.6536708860759495,
"grad_norm": 0.4075943503870774,
"learning_rate": 3.380718076670195e-05,
"loss": 0.3778,
"step": 408
},
{
"epoch": 1.6577215189873418,
"grad_norm": 0.32052029948824623,
"learning_rate": 3.376606219702454e-05,
"loss": 0.3761,
"step": 409
},
{
"epoch": 1.661772151898734,
"grad_norm": 0.4435598297145879,
"learning_rate": 3.372483275737468e-05,
"loss": 0.382,
"step": 410
},
{
"epoch": 1.6658227848101266,
"grad_norm": 0.29866062677885163,
"learning_rate": 3.368349277980861e-05,
"loss": 0.3738,
"step": 411
},
{
"epoch": 1.669873417721519,
"grad_norm": 0.38414341015755565,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.3771,
"step": 412
},
{
"epoch": 1.6739240506329114,
"grad_norm": 0.31013406093091356,
"learning_rate": 3.360048254360144e-05,
"loss": 0.3877,
"step": 413
},
{
"epoch": 1.6779746835443037,
"grad_norm": 0.39006319991405003,
"learning_rate": 3.355881295351336e-05,
"loss": 0.3658,
"step": 414
},
{
"epoch": 1.6820253164556962,
"grad_norm": 0.33905795551954254,
"learning_rate": 3.351703416260975e-05,
"loss": 0.3902,
"step": 415
},
{
"epoch": 1.6860759493670887,
"grad_norm": 0.36815935285420853,
"learning_rate": 3.347514650737126e-05,
"loss": 0.3882,
"step": 416
},
{
"epoch": 1.690126582278481,
"grad_norm": 0.41112591471991383,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.3823,
"step": 417
},
{
"epoch": 1.6941772151898733,
"grad_norm": 0.36811190378548725,
"learning_rate": 3.339104595419334e-05,
"loss": 0.3891,
"step": 418
},
{
"epoch": 1.6982278481012658,
"grad_norm": 0.3320714633258317,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.3887,
"step": 419
},
{
"epoch": 1.7022784810126583,
"grad_norm": 0.40507346764974056,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.3822,
"step": 420
},
{
"epoch": 1.7063291139240506,
"grad_norm": 0.42511181152724625,
"learning_rate": 3.326408710419996e-05,
"loss": 0.3818,
"step": 421
},
{
"epoch": 1.710379746835443,
"grad_norm": 0.3988967333155405,
"learning_rate": 3.322155337795454e-05,
"loss": 0.3837,
"step": 422
},
{
"epoch": 1.7144303797468354,
"grad_norm": 0.3460883447494473,
"learning_rate": 3.317891316713587e-05,
"loss": 0.3827,
"step": 423
},
{
"epoch": 1.718481012658228,
"grad_norm": 0.4061167943561236,
"learning_rate": 3.313616681516231e-05,
"loss": 0.3734,
"step": 424
},
{
"epoch": 1.7225316455696202,
"grad_norm": 0.4003115636236473,
"learning_rate": 3.309331466630713e-05,
"loss": 0.3887,
"step": 425
},
{
"epoch": 1.7265822784810125,
"grad_norm": 0.45231642494161606,
"learning_rate": 3.305035706569563e-05,
"loss": 0.3889,
"step": 426
},
{
"epoch": 1.730632911392405,
"grad_norm": 0.36171066327093815,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.3811,
"step": 427
},
{
"epoch": 1.7346835443037976,
"grad_norm": 0.4932635078404322,
"learning_rate": 3.296412689394864e-05,
"loss": 0.3778,
"step": 428
},
{
"epoch": 1.7387341772151899,
"grad_norm": 0.435459193717991,
"learning_rate": 3.292085501729909e-05,
"loss": 0.384,
"step": 429
},
{
"epoch": 1.7427848101265822,
"grad_norm": 0.39552089456297734,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.3777,
"step": 430
},
{
"epoch": 1.7468354430379747,
"grad_norm": 0.40714380626008917,
"learning_rate": 3.283399942497381e-05,
"loss": 0.3696,
"step": 431
},
{
"epoch": 1.7508860759493672,
"grad_norm": 0.3742613287868473,
"learning_rate": 3.279041640882108e-05,
"loss": 0.3901,
"step": 432
},
{
"epoch": 1.7549367088607595,
"grad_norm": 0.4560788803303788,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.3757,
"step": 433
},
{
"epoch": 1.7589873417721518,
"grad_norm": 0.387256403244439,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.3936,
"step": 434
},
{
"epoch": 1.7630379746835443,
"grad_norm": 0.3772726053055369,
"learning_rate": 3.265905069502244e-05,
"loss": 0.39,
"step": 435
},
{
"epoch": 1.7670886075949368,
"grad_norm": 0.31661332792439734,
"learning_rate": 3.261505774420016e-05,
"loss": 0.3727,
"step": 436
},
{
"epoch": 1.771139240506329,
"grad_norm": 0.35072487265832325,
"learning_rate": 3.257096319343707e-05,
"loss": 0.3835,
"step": 437
},
{
"epoch": 1.7751898734177214,
"grad_norm": 0.3968293904005149,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.387,
"step": 438
},
{
"epoch": 1.779240506329114,
"grad_norm": 0.40205024629424185,
"learning_rate": 3.248247071342966e-05,
"loss": 0.3933,
"step": 439
},
{
"epoch": 1.7832911392405064,
"grad_norm": 0.37504712430881,
"learning_rate": 3.243807349689161e-05,
"loss": 0.3791,
"step": 440
},
{
"epoch": 1.7873417721518987,
"grad_norm": 0.30876569592174763,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.3847,
"step": 441
},
{
"epoch": 1.791392405063291,
"grad_norm": 0.37300707810526545,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.3863,
"step": 442
},
{
"epoch": 1.7954430379746835,
"grad_norm": 0.37695679045317065,
"learning_rate": 3.230428223438075e-05,
"loss": 0.3836,
"step": 443
},
{
"epoch": 1.799493670886076,
"grad_norm": 0.36705645536904874,
"learning_rate": 3.225948647317459e-05,
"loss": 0.3838,
"step": 444
},
{
"epoch": 1.8035443037974683,
"grad_norm": 0.3057299823211106,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.3733,
"step": 445
},
{
"epoch": 1.8075949367088606,
"grad_norm": 0.40642770346910234,
"learning_rate": 3.216959910368034e-05,
"loss": 0.3778,
"step": 446
},
{
"epoch": 1.8116455696202531,
"grad_norm": 0.32650596138838056,
"learning_rate": 3.212450821933277e-05,
"loss": 0.3769,
"step": 447
},
{
"epoch": 1.8156962025316457,
"grad_norm": 0.3894078442423853,
"learning_rate": 3.207931968586281e-05,
"loss": 0.3863,
"step": 448
},
{
"epoch": 1.819746835443038,
"grad_norm": 0.35674177224173287,
"learning_rate": 3.203403386721272e-05,
"loss": 0.39,
"step": 449
},
{
"epoch": 1.8237974683544302,
"grad_norm": 0.3665659295335559,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.3906,
"step": 450
},
{
"epoch": 1.8278481012658228,
"grad_norm": 0.2968200088474544,
"learning_rate": 3.194317183405573e-05,
"loss": 0.382,
"step": 451
},
{
"epoch": 1.8318987341772153,
"grad_norm": 0.3678108636735864,
"learning_rate": 3.189759635133914e-05,
"loss": 0.3812,
"step": 452
},
{
"epoch": 1.8359493670886076,
"grad_norm": 0.3597283738290209,
"learning_rate": 3.185192504701718e-05,
"loss": 0.3754,
"step": 453
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.3165624678310885,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.384,
"step": 454
},
{
"epoch": 1.8440506329113924,
"grad_norm": 0.3196328368022559,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.3763,
"step": 455
},
{
"epoch": 1.8481012658227849,
"grad_norm": 0.293472657316469,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.3784,
"step": 456
},
{
"epoch": 1.8521518987341772,
"grad_norm": 0.2819334333342572,
"learning_rate": 3.166828898179731e-05,
"loss": 0.3684,
"step": 457
},
{
"epoch": 1.8562025316455695,
"grad_norm": 0.3179632373699564,
"learning_rate": 3.162214410223536e-05,
"loss": 0.3734,
"step": 458
},
{
"epoch": 1.860253164556962,
"grad_norm": 0.31331807371746445,
"learning_rate": 3.157590561952257e-05,
"loss": 0.3783,
"step": 459
},
{
"epoch": 1.8643037974683545,
"grad_norm": 0.2985256344146068,
"learning_rate": 3.152957390605732e-05,
"loss": 0.3842,
"step": 460
},
{
"epoch": 1.8683544303797468,
"grad_norm": 0.3512000293565593,
"learning_rate": 3.148314933498886e-05,
"loss": 0.3827,
"step": 461
},
{
"epoch": 1.872405063291139,
"grad_norm": 0.34802913282898185,
"learning_rate": 3.143663228021431e-05,
"loss": 0.3805,
"step": 462
},
{
"epoch": 1.8764556962025316,
"grad_norm": 0.2576219152249709,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.3745,
"step": 463
},
{
"epoch": 1.8805063291139241,
"grad_norm": 0.392446407039996,
"learning_rate": 3.134332221885661e-05,
"loss": 0.3903,
"step": 464
},
{
"epoch": 1.8845569620253164,
"grad_norm": 0.3663286762882459,
"learning_rate": 3.129652996377987e-05,
"loss": 0.369,
"step": 465
},
{
"epoch": 1.8886075949367087,
"grad_norm": 0.3213641594403502,
"learning_rate": 3.12496467280038e-05,
"loss": 0.3842,
"step": 466
},
{
"epoch": 1.8926582278481012,
"grad_norm": 0.35848741138475004,
"learning_rate": 3.120267288911952e-05,
"loss": 0.3758,
"step": 467
},
{
"epoch": 1.8967088607594937,
"grad_norm": 0.38583424997340643,
"learning_rate": 3.11556088254479e-05,
"loss": 0.3714,
"step": 468
},
{
"epoch": 1.900759493670886,
"grad_norm": 0.3240424172697002,
"learning_rate": 3.11084549160364e-05,
"loss": 0.382,
"step": 469
},
{
"epoch": 1.9048101265822783,
"grad_norm": 0.3488382941618938,
"learning_rate": 3.106121154065615e-05,
"loss": 0.3751,
"step": 470
},
{
"epoch": 1.9088607594936708,
"grad_norm": 0.3699090108825789,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.3754,
"step": 471
},
{
"epoch": 1.9129113924050634,
"grad_norm": 0.3365654662673872,
"learning_rate": 3.096645791467348e-05,
"loss": 0.3734,
"step": 472
},
{
"epoch": 1.9169620253164557,
"grad_norm": 0.3388468121382902,
"learning_rate": 3.091894842720373e-05,
"loss": 0.3846,
"step": 473
},
{
"epoch": 1.921012658227848,
"grad_norm": 0.3442816933858571,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.3753,
"step": 474
},
{
"epoch": 1.9250632911392405,
"grad_norm": 0.398567960499571,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.3818,
"step": 475
},
{
"epoch": 1.929113924050633,
"grad_norm": 0.31015598249584525,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.3752,
"step": 476
},
{
"epoch": 1.9331645569620253,
"grad_norm": 0.3906500814258093,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.3829,
"step": 477
},
{
"epoch": 1.9372151898734176,
"grad_norm": 0.33255703756998245,
"learning_rate": 3.06800895716334e-05,
"loss": 0.3784,
"step": 478
},
{
"epoch": 1.94126582278481,
"grad_norm": 0.31811501268098946,
"learning_rate": 3.063205821011029e-05,
"loss": 0.3683,
"step": 479
},
{
"epoch": 1.9453164556962026,
"grad_norm": 0.3149525935530576,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.3753,
"step": 480
},
{
"epoch": 1.9493670886075949,
"grad_norm": 0.29766572576536754,
"learning_rate": 3.053573898718618e-05,
"loss": 0.3792,
"step": 481
},
{
"epoch": 1.9534177215189872,
"grad_norm": 0.3225381646592208,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.3663,
"step": 482
},
{
"epoch": 1.9574683544303797,
"grad_norm": 0.2978170309578759,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.387,
"step": 483
},
{
"epoch": 1.9615189873417722,
"grad_norm": 0.3451585096707341,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.3804,
"step": 484
},
{
"epoch": 1.9655696202531645,
"grad_norm": 0.34536924565263716,
"learning_rate": 3.034208541656255e-05,
"loss": 0.3668,
"step": 485
},
{
"epoch": 1.9696202531645568,
"grad_norm": 0.3508669401822276,
"learning_rate": 3.029346281309218e-05,
"loss": 0.378,
"step": 486
},
{
"epoch": 1.9736708860759493,
"grad_norm": 0.3862483599481415,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.3907,
"step": 487
},
{
"epoch": 1.9777215189873418,
"grad_norm": 0.28380134489636516,
"learning_rate": 3.019596929201569e-05,
"loss": 0.3741,
"step": 488
},
{
"epoch": 1.9817721518987341,
"grad_norm": 0.4074958261993291,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.3741,
"step": 489
},
{
"epoch": 1.9858227848101264,
"grad_norm": 0.28264162632471584,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.3829,
"step": 490
},
{
"epoch": 1.989873417721519,
"grad_norm": 0.35294352408085894,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.3783,
"step": 491
},
{
"epoch": 1.9939240506329114,
"grad_norm": 0.28758829210648407,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.3676,
"step": 492
},
{
"epoch": 1.9979746835443037,
"grad_norm": 0.32945830406379467,
"learning_rate": 2.995080534235264e-05,
"loss": 0.3839,
"step": 493
},
{
"epoch": 2.0030379746835445,
"grad_norm": 0.34091361182658086,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.343,
"step": 494
},
{
"epoch": 2.0070886075949366,
"grad_norm": 0.30188954308001953,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3119,
"step": 495
},
{
"epoch": 2.011139240506329,
"grad_norm": 0.5100078747548665,
"learning_rate": 2.980274210295326e-05,
"loss": 0.3256,
"step": 496
},
{
"epoch": 2.0151898734177216,
"grad_norm": 0.3699024173315946,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.3299,
"step": 497
},
{
"epoch": 2.019240506329114,
"grad_norm": 0.4115991236775706,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.3414,
"step": 498
},
{
"epoch": 2.023291139240506,
"grad_norm": 0.3343608528893723,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3208,
"step": 499
},
{
"epoch": 2.0273417721518987,
"grad_norm": 0.3872388382326177,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.3325,
"step": 500
},
{
"epoch": 2.031392405063291,
"grad_norm": 0.3780497286768413,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3284,
"step": 501
},
{
"epoch": 2.0354430379746837,
"grad_norm": 0.35241059411930925,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3371,
"step": 502
},
{
"epoch": 2.039493670886076,
"grad_norm": 0.3812839109993378,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.3368,
"step": 503
},
{
"epoch": 2.0435443037974683,
"grad_norm": 0.29623919216001815,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3336,
"step": 504
},
{
"epoch": 2.047594936708861,
"grad_norm": 0.3390674662693876,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3265,
"step": 505
},
{
"epoch": 2.0516455696202534,
"grad_norm": 0.31261765470882524,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.3229,
"step": 506
},
{
"epoch": 2.0556962025316454,
"grad_norm": 0.33063211822190686,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.326,
"step": 507
},
{
"epoch": 2.059746835443038,
"grad_norm": 0.3089399434898628,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3312,
"step": 508
},
{
"epoch": 2.0637974683544305,
"grad_norm": 0.3742807150754224,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.3295,
"step": 509
},
{
"epoch": 2.067848101265823,
"grad_norm": 0.2928190610873682,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.3219,
"step": 510
},
{
"epoch": 2.071898734177215,
"grad_norm": 0.3193491014406378,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3196,
"step": 511
},
{
"epoch": 2.0759493670886076,
"grad_norm": 0.3433466740415105,
"learning_rate": 2.900133526131588e-05,
"loss": 0.3252,
"step": 512
},
{
"epoch": 2.08,
"grad_norm": 0.3753623492896508,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.3398,
"step": 513
},
{
"epoch": 2.0840506329113926,
"grad_norm": 0.3389455585459374,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3354,
"step": 514
},
{
"epoch": 2.0881012658227847,
"grad_norm": 0.39625634373892304,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3321,
"step": 515
},
{
"epoch": 2.092151898734177,
"grad_norm": 0.38470931714319534,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3206,
"step": 516
},
{
"epoch": 2.0962025316455697,
"grad_norm": 0.34739998609179024,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.3245,
"step": 517
},
{
"epoch": 2.100253164556962,
"grad_norm": 0.4052398782930149,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.3311,
"step": 518
},
{
"epoch": 2.1043037974683543,
"grad_norm": 0.404831233470577,
"learning_rate": 2.864478605989494e-05,
"loss": 0.3314,
"step": 519
},
{
"epoch": 2.108354430379747,
"grad_norm": 0.3772370285642263,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.3304,
"step": 520
},
{
"epoch": 2.1124050632911393,
"grad_norm": 0.33164557566159997,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3204,
"step": 521
},
{
"epoch": 2.116455696202532,
"grad_norm": 0.34603408047679174,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3256,
"step": 522
},
{
"epoch": 2.120506329113924,
"grad_norm": 0.3537261938974695,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.3236,
"step": 523
},
{
"epoch": 2.1245569620253164,
"grad_norm": 0.33386831597664635,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.3195,
"step": 524
},
{
"epoch": 2.128607594936709,
"grad_norm": 0.3935948554246293,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3291,
"step": 525
},
{
"epoch": 2.1326582278481014,
"grad_norm": 0.4321183266056485,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.3296,
"step": 526
},
{
"epoch": 2.1367088607594935,
"grad_norm": 0.3610132906681355,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.338,
"step": 527
},
{
"epoch": 2.140759493670886,
"grad_norm": 0.40253480549533605,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.3394,
"step": 528
},
{
"epoch": 2.1448101265822785,
"grad_norm": 0.3353134064397972,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3246,
"step": 529
},
{
"epoch": 2.148860759493671,
"grad_norm": 0.41325739237979553,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3252,
"step": 530
},
{
"epoch": 2.152911392405063,
"grad_norm": 0.33630888311515944,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3232,
"step": 531
},
{
"epoch": 2.1569620253164556,
"grad_norm": 0.33727168612499625,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3306,
"step": 532
},
{
"epoch": 2.161012658227848,
"grad_norm": 0.3191645483485916,
"learning_rate": 2.792159532078314e-05,
"loss": 0.33,
"step": 533
},
{
"epoch": 2.1650632911392407,
"grad_norm": 0.32545642428146204,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.3192,
"step": 534
},
{
"epoch": 2.1691139240506327,
"grad_norm": 0.33905150329790645,
"learning_rate": 2.781723482288857e-05,
"loss": 0.333,
"step": 535
},
{
"epoch": 2.1731645569620253,
"grad_norm": 0.2922936444596804,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.3351,
"step": 536
},
{
"epoch": 2.1772151898734178,
"grad_norm": 0.30295456547918653,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3261,
"step": 537
},
{
"epoch": 2.1812658227848103,
"grad_norm": 0.31934708809185175,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.3371,
"step": 538
},
{
"epoch": 2.1853164556962024,
"grad_norm": 0.3257153708781759,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3325,
"step": 539
},
{
"epoch": 2.189367088607595,
"grad_norm": 0.3359072194439771,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.3385,
"step": 540
},
{
"epoch": 2.1934177215189874,
"grad_norm": 0.33335386263433,
"learning_rate": 2.7502655806328e-05,
"loss": 0.328,
"step": 541
},
{
"epoch": 2.19746835443038,
"grad_norm": 0.3330524446490538,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3332,
"step": 542
},
{
"epoch": 2.201518987341772,
"grad_norm": 0.3079098616894255,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3178,
"step": 543
},
{
"epoch": 2.2055696202531645,
"grad_norm": 0.36533581929646053,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.3278,
"step": 544
},
{
"epoch": 2.209620253164557,
"grad_norm": 0.4005273116424787,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.3366,
"step": 545
},
{
"epoch": 2.2136708860759495,
"grad_norm": 0.3122379504321191,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3339,
"step": 546
},
{
"epoch": 2.2177215189873416,
"grad_norm": 0.3951112667397085,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3249,
"step": 547
},
{
"epoch": 2.221772151898734,
"grad_norm": 0.33112915968775014,
"learning_rate": 2.71329041039194e-05,
"loss": 0.336,
"step": 548
},
{
"epoch": 2.2258227848101266,
"grad_norm": 0.3015305183994437,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3334,
"step": 549
},
{
"epoch": 2.229873417721519,
"grad_norm": 0.31901855957317865,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3161,
"step": 550
},
{
"epoch": 2.233924050632911,
"grad_norm": 0.320996490996312,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3381,
"step": 551
},
{
"epoch": 2.2379746835443037,
"grad_norm": 0.32548361724757036,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3307,
"step": 552
},
{
"epoch": 2.2420253164556962,
"grad_norm": 0.298276653585191,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.3263,
"step": 553
},
{
"epoch": 2.2460759493670888,
"grad_norm": 0.31994314648679,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.3321,
"step": 554
},
{
"epoch": 2.250126582278481,
"grad_norm": 0.28613327733097543,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3153,
"step": 555
},
{
"epoch": 2.2541772151898734,
"grad_norm": 0.31160215142176245,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.3387,
"step": 556
},
{
"epoch": 2.258227848101266,
"grad_norm": 0.31619315767163053,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3285,
"step": 557
},
{
"epoch": 2.2622784810126584,
"grad_norm": 0.2544737485705707,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.3276,
"step": 558
},
{
"epoch": 2.2663291139240505,
"grad_norm": 0.35046075455297754,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.3357,
"step": 559
},
{
"epoch": 2.270379746835443,
"grad_norm": 0.27354538368580583,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.333,
"step": 560
},
{
"epoch": 2.2744303797468355,
"grad_norm": 0.30888961821935335,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.3307,
"step": 561
},
{
"epoch": 2.278481012658228,
"grad_norm": 0.31819201387482904,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.331,
"step": 562
},
{
"epoch": 2.28253164556962,
"grad_norm": 0.33191373816461817,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.336,
"step": 563
},
{
"epoch": 2.2865822784810126,
"grad_norm": 0.34074460868095147,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.3299,
"step": 564
},
{
"epoch": 2.290632911392405,
"grad_norm": 0.27257019912458297,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3262,
"step": 565
},
{
"epoch": 2.2946835443037976,
"grad_norm": 0.3286678754569777,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.331,
"step": 566
},
{
"epoch": 2.2987341772151897,
"grad_norm": 0.306345834715826,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3233,
"step": 567
},
{
"epoch": 2.302784810126582,
"grad_norm": 0.3170260970584045,
"learning_rate": 2.606146264569603e-05,
"loss": 0.3255,
"step": 568
},
{
"epoch": 2.3068354430379747,
"grad_norm": 0.3256515118651009,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3242,
"step": 569
},
{
"epoch": 2.3108860759493672,
"grad_norm": 0.28249418018307154,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.3353,
"step": 570
},
{
"epoch": 2.3149367088607593,
"grad_norm": 0.3270675674506942,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.3245,
"step": 571
},
{
"epoch": 2.318987341772152,
"grad_norm": 0.2665354485260207,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3335,
"step": 572
},
{
"epoch": 2.3230379746835443,
"grad_norm": 0.3641729483500324,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3334,
"step": 573
},
{
"epoch": 2.327088607594937,
"grad_norm": 0.25329325197990094,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3219,
"step": 574
},
{
"epoch": 2.331139240506329,
"grad_norm": 0.3614521858019154,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3309,
"step": 575
},
{
"epoch": 2.3351898734177214,
"grad_norm": 0.2673053381583691,
"learning_rate": 2.56272244128509e-05,
"loss": 0.3252,
"step": 576
},
{
"epoch": 2.339240506329114,
"grad_norm": 0.29973264678721684,
"learning_rate": 2.55727360915895e-05,
"loss": 0.3231,
"step": 577
},
{
"epoch": 2.3432911392405065,
"grad_norm": 0.2603627870703053,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3333,
"step": 578
},
{
"epoch": 2.3473417721518985,
"grad_norm": 0.30584279847544277,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3377,
"step": 579
},
{
"epoch": 2.351392405063291,
"grad_norm": 0.2468120425855416,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3298,
"step": 580
},
{
"epoch": 2.3554430379746836,
"grad_norm": 0.29454636522768357,
"learning_rate": 2.535433837985559e-05,
"loss": 0.3403,
"step": 581
},
{
"epoch": 2.359493670886076,
"grad_norm": 0.2434978616959565,
"learning_rate": 2.529963004393324e-05,
"loss": 0.3219,
"step": 582
},
{
"epoch": 2.363544303797468,
"grad_norm": 0.314493512168681,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3262,
"step": 583
},
{
"epoch": 2.3675949367088607,
"grad_norm": 0.2515330035317457,
"learning_rate": 2.519008576557029e-05,
"loss": 0.3246,
"step": 584
},
{
"epoch": 2.371645569620253,
"grad_norm": 0.28796799118719196,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.3329,
"step": 585
},
{
"epoch": 2.3756962025316457,
"grad_norm": 0.26626804192021325,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.3195,
"step": 586
},
{
"epoch": 2.379746835443038,
"grad_norm": 0.2759498539960967,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.3202,
"step": 587
},
{
"epoch": 2.3837974683544303,
"grad_norm": 0.29269480899692263,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3337,
"step": 588
},
{
"epoch": 2.387848101265823,
"grad_norm": 0.28304258287554696,
"learning_rate": 2.491550129992345e-05,
"loss": 0.336,
"step": 589
},
{
"epoch": 2.3918987341772153,
"grad_norm": 0.3142057630880705,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3288,
"step": 590
},
{
"epoch": 2.3959493670886074,
"grad_norm": 0.27361686729613266,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.328,
"step": 591
},
{
"epoch": 2.4,
"grad_norm": 0.33917796196766703,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.3269,
"step": 592
},
{
"epoch": 2.4040506329113924,
"grad_norm": 0.2952899919114709,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3329,
"step": 593
},
{
"epoch": 2.408101265822785,
"grad_norm": 0.34420581664672834,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.3296,
"step": 594
},
{
"epoch": 2.4121518987341775,
"grad_norm": 0.28856913239893256,
"learning_rate": 2.458469841202444e-05,
"loss": 0.3313,
"step": 595
},
{
"epoch": 2.4162025316455695,
"grad_norm": 0.35074466339955285,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3322,
"step": 596
},
{
"epoch": 2.420253164556962,
"grad_norm": 0.286995176145559,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3401,
"step": 597
},
{
"epoch": 2.4243037974683546,
"grad_norm": 0.32254753114362805,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.3202,
"step": 598
},
{
"epoch": 2.4283544303797466,
"grad_norm": 0.31634769575559013,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.3237,
"step": 599
},
{
"epoch": 2.432405063291139,
"grad_norm": 0.28910091701989343,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3237,
"step": 600
},
{
"epoch": 2.4364556962025317,
"grad_norm": 0.30486690839517294,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3225,
"step": 601
},
{
"epoch": 2.440506329113924,
"grad_norm": 0.28547239853789486,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3336,
"step": 602
},
{
"epoch": 2.4445569620253167,
"grad_norm": 0.35040921150822174,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.328,
"step": 603
},
{
"epoch": 2.4486075949367088,
"grad_norm": 0.2959408959577134,
"learning_rate": 2.408603181104957e-05,
"loss": 0.3189,
"step": 604
},
{
"epoch": 2.4526582278481013,
"grad_norm": 0.29957024995027093,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.336,
"step": 605
},
{
"epoch": 2.456708860759494,
"grad_norm": 0.34007469922531,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3322,
"step": 606
},
{
"epoch": 2.460759493670886,
"grad_norm": 0.28549250661978026,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.3267,
"step": 607
},
{
"epoch": 2.4648101265822784,
"grad_norm": 0.35429147727498006,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.3246,
"step": 608
},
{
"epoch": 2.468860759493671,
"grad_norm": 0.3148679186421029,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3409,
"step": 609
},
{
"epoch": 2.4729113924050634,
"grad_norm": 0.31723637458202425,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.3268,
"step": 610
},
{
"epoch": 2.476962025316456,
"grad_norm": 0.2882992478160064,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3213,
"step": 611
},
{
"epoch": 2.481012658227848,
"grad_norm": 0.26405684424207876,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.3221,
"step": 612
},
{
"epoch": 2.4850632911392405,
"grad_norm": 0.29992393813592244,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3331,
"step": 613
},
{
"epoch": 2.489113924050633,
"grad_norm": 0.29504666440077865,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3363,
"step": 614
},
{
"epoch": 2.493164556962025,
"grad_norm": 0.31983583043648606,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.3334,
"step": 615
},
{
"epoch": 2.4972151898734176,
"grad_norm": 0.2863386628304591,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3335,
"step": 616
},
{
"epoch": 2.50126582278481,
"grad_norm": 0.26765839026612076,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3325,
"step": 617
},
{
"epoch": 2.5053164556962026,
"grad_norm": 0.2678870943795969,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3258,
"step": 618
},
{
"epoch": 2.509367088607595,
"grad_norm": 0.24764743210567183,
"learning_rate": 2.324915905911693e-05,
"loss": 0.3202,
"step": 619
},
{
"epoch": 2.5134177215189872,
"grad_norm": 0.2725801642293375,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3351,
"step": 620
},
{
"epoch": 2.5174683544303798,
"grad_norm": 0.26285367301740814,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.334,
"step": 621
},
{
"epoch": 2.5215189873417723,
"grad_norm": 0.2443528310214279,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3181,
"step": 622
},
{
"epoch": 2.5255696202531643,
"grad_norm": 0.26673290695539614,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.331,
"step": 623
},
{
"epoch": 2.529620253164557,
"grad_norm": 0.2562546489200084,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3285,
"step": 624
},
{
"epoch": 2.5336708860759494,
"grad_norm": 0.32421855738004424,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3198,
"step": 625
},
{
"epoch": 2.537721518987342,
"grad_norm": 0.27261340670289214,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3374,
"step": 626
},
{
"epoch": 2.5417721518987344,
"grad_norm": 0.3211722655824439,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3401,
"step": 627
},
{
"epoch": 2.5458227848101265,
"grad_norm": 0.2756728407589125,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.3232,
"step": 628
},
{
"epoch": 2.549873417721519,
"grad_norm": 0.32614700140471203,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.339,
"step": 629
},
{
"epoch": 2.5539240506329115,
"grad_norm": 0.3175657035119466,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3342,
"step": 630
},
{
"epoch": 2.5579746835443036,
"grad_norm": 0.3431558163037973,
"learning_rate": 2.257534907110328e-05,
"loss": 0.34,
"step": 631
},
{
"epoch": 2.562025316455696,
"grad_norm": 0.2588689936101726,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3289,
"step": 632
},
{
"epoch": 2.5660759493670886,
"grad_norm": 0.34476253021044473,
"learning_rate": 2.246273589557294e-05,
"loss": 0.3212,
"step": 633
},
{
"epoch": 2.570126582278481,
"grad_norm": 0.2752165095107865,
"learning_rate": 2.240639932921966e-05,
"loss": 0.333,
"step": 634
},
{
"epoch": 2.5741772151898736,
"grad_norm": 0.298271031355723,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.3139,
"step": 635
},
{
"epoch": 2.5782278481012657,
"grad_norm": 0.26814744621126374,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.3227,
"step": 636
},
{
"epoch": 2.5822784810126582,
"grad_norm": 0.2738540379467409,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.3348,
"step": 637
},
{
"epoch": 2.5863291139240507,
"grad_norm": 0.28276321445484814,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.3213,
"step": 638
},
{
"epoch": 2.590379746835443,
"grad_norm": 0.2824464181505588,
"learning_rate": 2.212443486523819e-05,
"loss": 0.334,
"step": 639
},
{
"epoch": 2.5944303797468353,
"grad_norm": 0.2907181545933403,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.3225,
"step": 640
},
{
"epoch": 2.598481012658228,
"grad_norm": 0.2692541343321853,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3404,
"step": 641
},
{
"epoch": 2.6025316455696204,
"grad_norm": 0.26212050973809714,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3194,
"step": 642
},
{
"epoch": 2.606582278481013,
"grad_norm": 0.29806571078604693,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3321,
"step": 643
},
{
"epoch": 2.610632911392405,
"grad_norm": 0.3026003009607464,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3183,
"step": 644
},
{
"epoch": 2.6146835443037975,
"grad_norm": 0.2668042471765998,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.3202,
"step": 645
},
{
"epoch": 2.61873417721519,
"grad_norm": 0.3059549514062728,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3292,
"step": 646
},
{
"epoch": 2.622784810126582,
"grad_norm": 0.26853030005408035,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.33,
"step": 647
},
{
"epoch": 2.6268354430379746,
"grad_norm": 0.25557550076332847,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3276,
"step": 648
},
{
"epoch": 2.630886075949367,
"grad_norm": 0.24840650747468584,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.321,
"step": 649
},
{
"epoch": 2.6349367088607596,
"grad_norm": 0.2642100386513847,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3228,
"step": 650
},
{
"epoch": 2.638987341772152,
"grad_norm": 0.296627999376536,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.3354,
"step": 651
},
{
"epoch": 2.643037974683544,
"grad_norm": 0.3074937120133868,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3328,
"step": 652
},
{
"epoch": 2.6470886075949367,
"grad_norm": 0.2388748608865812,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3205,
"step": 653
},
{
"epoch": 2.651139240506329,
"grad_norm": 0.29327777981907427,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3369,
"step": 654
},
{
"epoch": 2.6551898734177213,
"grad_norm": 0.2639405232922625,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3294,
"step": 655
},
{
"epoch": 2.659240506329114,
"grad_norm": 0.2638844719408201,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3265,
"step": 656
},
{
"epoch": 2.6632911392405063,
"grad_norm": 0.26727552615194766,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.3293,
"step": 657
},
{
"epoch": 2.667341772151899,
"grad_norm": 0.24573046427250594,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3214,
"step": 658
},
{
"epoch": 2.6713924050632913,
"grad_norm": 0.24440601374422358,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3353,
"step": 659
},
{
"epoch": 2.6754430379746834,
"grad_norm": 0.2354925069140139,
"learning_rate": 2.093617592756158e-05,
"loss": 0.332,
"step": 660
},
{
"epoch": 2.679493670886076,
"grad_norm": 0.23734300094321284,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3216,
"step": 661
},
{
"epoch": 2.6835443037974684,
"grad_norm": 0.23911193035985728,
"learning_rate": 2.082276853708978e-05,
"loss": 0.3219,
"step": 662
},
{
"epoch": 2.6875949367088605,
"grad_norm": 0.2206897954759037,
"learning_rate": 2.076605467380071e-05,
"loss": 0.334,
"step": 663
},
{
"epoch": 2.691645569620253,
"grad_norm": 0.2635783693815686,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3307,
"step": 664
},
{
"epoch": 2.6956962025316455,
"grad_norm": 0.24532967662702054,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3346,
"step": 665
},
{
"epoch": 2.699746835443038,
"grad_norm": 0.2629478681525093,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3434,
"step": 666
},
{
"epoch": 2.7037974683544306,
"grad_norm": 0.2816672199415327,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.3378,
"step": 667
},
{
"epoch": 2.7078481012658226,
"grad_norm": 0.26884277382903404,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3164,
"step": 668
},
{
"epoch": 2.711898734177215,
"grad_norm": 0.2656600915484324,
"learning_rate": 2.042565792045695e-05,
"loss": 0.3204,
"step": 669
},
{
"epoch": 2.7159493670886077,
"grad_norm": 0.25206161455674664,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3179,
"step": 670
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.24654321763311407,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.3202,
"step": 671
},
{
"epoch": 2.7240506329113923,
"grad_norm": 0.2535578133649192,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3236,
"step": 672
},
{
"epoch": 2.728101265822785,
"grad_norm": 0.2849374104685301,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3346,
"step": 673
},
{
"epoch": 2.7321518987341773,
"grad_norm": 0.26121887822179757,
"learning_rate": 2.014189549672245e-05,
"loss": 0.329,
"step": 674
},
{
"epoch": 2.73620253164557,
"grad_norm": 0.2812974258824899,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3304,
"step": 675
},
{
"epoch": 2.740253164556962,
"grad_norm": 0.26406007594315317,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3256,
"step": 676
},
{
"epoch": 2.7443037974683544,
"grad_norm": 0.29010114338041365,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.3267,
"step": 677
},
{
"epoch": 2.748354430379747,
"grad_norm": 0.280856262140703,
"learning_rate": 1.991486224484123e-05,
"loss": 0.3383,
"step": 678
},
{
"epoch": 2.752405063291139,
"grad_norm": 0.3132133071242575,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3173,
"step": 679
},
{
"epoch": 2.7564556962025315,
"grad_norm": 0.26285350992405215,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.3268,
"step": 680
},
{
"epoch": 2.760506329113924,
"grad_norm": 0.3099591068165603,
"learning_rate": 1.974459290568053e-05,
"loss": 0.3192,
"step": 681
},
{
"epoch": 2.7645569620253165,
"grad_norm": 0.2800885457385436,
"learning_rate": 1.968783996385397e-05,
"loss": 0.3156,
"step": 682
},
{
"epoch": 2.768607594936709,
"grad_norm": 0.29432930472537366,
"learning_rate": 1.963108953612143e-05,
"loss": 0.3348,
"step": 683
},
{
"epoch": 2.772658227848101,
"grad_norm": 0.2759429081247148,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3273,
"step": 684
},
{
"epoch": 2.7767088607594936,
"grad_norm": 0.24636521289574514,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3349,
"step": 685
},
{
"epoch": 2.780759493670886,
"grad_norm": 0.31309810393472876,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.3368,
"step": 686
},
{
"epoch": 2.7848101265822782,
"grad_norm": 0.22857588339240967,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3285,
"step": 687
},
{
"epoch": 2.7888607594936707,
"grad_norm": 0.3072766455441824,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.327,
"step": 688
},
{
"epoch": 2.7929113924050633,
"grad_norm": 0.22565221276672265,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3294,
"step": 689
},
{
"epoch": 2.7969620253164558,
"grad_norm": 0.33127486238979353,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.3279,
"step": 690
},
{
"epoch": 2.8010126582278483,
"grad_norm": 0.25960108744236726,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3209,
"step": 691
},
{
"epoch": 2.8050632911392404,
"grad_norm": 0.3409916125426889,
"learning_rate": 1.912052422608584e-05,
"loss": 0.3195,
"step": 692
},
{
"epoch": 2.809113924050633,
"grad_norm": 0.28183680494957475,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.3386,
"step": 693
},
{
"epoch": 2.8131645569620254,
"grad_norm": 0.3015017158596458,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3273,
"step": 694
},
{
"epoch": 2.8172151898734175,
"grad_norm": 0.25771143489281506,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3236,
"step": 695
},
{
"epoch": 2.82126582278481,
"grad_norm": 0.29160191967572346,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3452,
"step": 696
},
{
"epoch": 2.8253164556962025,
"grad_norm": 0.2559707852830831,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3262,
"step": 697
},
{
"epoch": 2.829367088607595,
"grad_norm": 0.2568308846851837,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3296,
"step": 698
},
{
"epoch": 2.8334177215189875,
"grad_norm": 0.27928075453509427,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3307,
"step": 699
},
{
"epoch": 2.8374683544303796,
"grad_norm": 0.22020299179838615,
"learning_rate": 1.866715967678769e-05,
"loss": 0.3175,
"step": 700
},
{
"epoch": 2.841518987341772,
"grad_norm": 0.24833896500265137,
"learning_rate": 1.861053262258067e-05,
"loss": 0.329,
"step": 701
},
{
"epoch": 2.8455696202531646,
"grad_norm": 0.2602231104941992,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3308,
"step": 702
},
{
"epoch": 2.8496202531645567,
"grad_norm": 0.2308273396852072,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3249,
"step": 703
},
{
"epoch": 2.853670886075949,
"grad_norm": 0.26201917728849033,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3269,
"step": 704
},
{
"epoch": 2.8577215189873417,
"grad_norm": 0.2542387572935997,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.3276,
"step": 705
},
{
"epoch": 2.8617721518987342,
"grad_norm": 0.2426331734852078,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3315,
"step": 706
},
{
"epoch": 2.8658227848101268,
"grad_norm": 0.24729986934318318,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.3249,
"step": 707
},
{
"epoch": 2.869873417721519,
"grad_norm": 0.27418004524445877,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.336,
"step": 708
},
{
"epoch": 2.8739240506329113,
"grad_norm": 0.2992544553206878,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3315,
"step": 709
},
{
"epoch": 2.877974683544304,
"grad_norm": 0.23881455053421963,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.3128,
"step": 710
},
{
"epoch": 2.882025316455696,
"grad_norm": 0.3020492134608753,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.321,
"step": 711
},
{
"epoch": 2.8860759493670884,
"grad_norm": 0.23402586091689478,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.3278,
"step": 712
},
{
"epoch": 2.890126582278481,
"grad_norm": 0.2391914733517342,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3276,
"step": 713
},
{
"epoch": 2.8941772151898735,
"grad_norm": 0.2735845197946949,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.3271,
"step": 714
},
{
"epoch": 2.898227848101266,
"grad_norm": 0.2336998455032747,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.3231,
"step": 715
},
{
"epoch": 2.902278481012658,
"grad_norm": 0.26922542745415085,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3329,
"step": 716
},
{
"epoch": 2.9063291139240506,
"grad_norm": 0.2788088405877167,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.3246,
"step": 717
},
{
"epoch": 2.910379746835443,
"grad_norm": 0.2691512383255931,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3295,
"step": 718
},
{
"epoch": 2.9144303797468356,
"grad_norm": 0.27236352977128764,
"learning_rate": 1.759360067078035e-05,
"loss": 0.3244,
"step": 719
},
{
"epoch": 2.918481012658228,
"grad_norm": 0.2486560551668802,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3199,
"step": 720
},
{
"epoch": 2.92253164556962,
"grad_norm": 0.2526215192288239,
"learning_rate": 1.748094737261012e-05,
"loss": 0.3172,
"step": 721
},
{
"epoch": 2.9265822784810127,
"grad_norm": 0.24233470129254717,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3277,
"step": 722
},
{
"epoch": 2.9306329113924052,
"grad_norm": 0.25691124861928566,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.317,
"step": 723
},
{
"epoch": 2.9346835443037973,
"grad_norm": 0.2648152221221479,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3287,
"step": 724
},
{
"epoch": 2.93873417721519,
"grad_norm": 0.24112550598282437,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.324,
"step": 725
},
{
"epoch": 2.9427848101265823,
"grad_norm": 0.2664361484881387,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.3397,
"step": 726
},
{
"epoch": 2.946835443037975,
"grad_norm": 0.2324588631840948,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3317,
"step": 727
},
{
"epoch": 2.9508860759493674,
"grad_norm": 0.28140913437399395,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.3301,
"step": 728
},
{
"epoch": 2.9549367088607594,
"grad_norm": 0.2374277466434981,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3353,
"step": 729
},
{
"epoch": 2.958987341772152,
"grad_norm": 0.24982285445755956,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3201,
"step": 730
},
{
"epoch": 2.9630379746835445,
"grad_norm": 0.25292380001625986,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3302,
"step": 731
},
{
"epoch": 2.9670886075949365,
"grad_norm": 0.2371348489118273,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3352,
"step": 732
},
{
"epoch": 2.971139240506329,
"grad_norm": 0.21026229326365456,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3297,
"step": 733
},
{
"epoch": 2.9751898734177216,
"grad_norm": 0.22796587802141188,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3286,
"step": 734
},
{
"epoch": 2.979240506329114,
"grad_norm": 0.2614410387191667,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3201,
"step": 735
},
{
"epoch": 2.9832911392405066,
"grad_norm": 0.2325773387430153,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3258,
"step": 736
},
{
"epoch": 2.9873417721518987,
"grad_norm": 0.2600901814798085,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3342,
"step": 737
},
{
"epoch": 2.991392405063291,
"grad_norm": 0.22968499941412288,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3247,
"step": 738
},
{
"epoch": 2.9954430379746837,
"grad_norm": 0.22189223087141643,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3111,
"step": 739
},
{
"epoch": 3.0005063291139242,
"grad_norm": 0.26745815964742103,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.3267,
"step": 740
},
{
"epoch": 3.0045569620253163,
"grad_norm": 0.4045857878940443,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.2806,
"step": 741
},
{
"epoch": 3.008607594936709,
"grad_norm": 0.4237238553114083,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2843,
"step": 742
},
{
"epoch": 3.0126582278481013,
"grad_norm": 0.3681242426978064,
"learning_rate": 1.624791353909428e-05,
"loss": 0.2734,
"step": 743
},
{
"epoch": 3.016708860759494,
"grad_norm": 0.4281153523757835,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2897,
"step": 744
},
{
"epoch": 3.020759493670886,
"grad_norm": 0.3445416404732803,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.279,
"step": 745
},
{
"epoch": 3.0248101265822784,
"grad_norm": 0.3194579336965379,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2639,
"step": 746
},
{
"epoch": 3.028860759493671,
"grad_norm": 0.33613782829824307,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2777,
"step": 747
},
{
"epoch": 3.0329113924050635,
"grad_norm": 0.31528227024828426,
"learning_rate": 1.596954608907176e-05,
"loss": 0.2778,
"step": 748
},
{
"epoch": 3.0369620253164555,
"grad_norm": 0.27336366063383033,
"learning_rate": 1.591396818895043e-05,
"loss": 0.2841,
"step": 749
},
{
"epoch": 3.041012658227848,
"grad_norm": 0.3171357990582387,
"learning_rate": 1.585842319716759e-05,
"loss": 0.2806,
"step": 750
},
{
"epoch": 3.0450632911392406,
"grad_norm": 0.2897463939098347,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2727,
"step": 751
},
{
"epoch": 3.049113924050633,
"grad_norm": 0.27236794739813636,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2757,
"step": 752
},
{
"epoch": 3.053164556962025,
"grad_norm": 0.31509556059763816,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.285,
"step": 753
},
{
"epoch": 3.0572151898734177,
"grad_norm": 0.27896852148521545,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2767,
"step": 754
},
{
"epoch": 3.06126582278481,
"grad_norm": 0.2699705786426204,
"learning_rate": 1.558120751116291e-05,
"loss": 0.2785,
"step": 755
},
{
"epoch": 3.0653164556962027,
"grad_norm": 0.26404869099547534,
"learning_rate": 1.552586935426876e-05,
"loss": 0.2649,
"step": 756
},
{
"epoch": 3.0693670886075948,
"grad_norm": 0.3027789913530549,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2696,
"step": 757
},
{
"epoch": 3.0734177215189873,
"grad_norm": 0.23452460659605248,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.2726,
"step": 758
},
{
"epoch": 3.07746835443038,
"grad_norm": 0.2661119343867333,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2817,
"step": 759
},
{
"epoch": 3.0815189873417723,
"grad_norm": 0.2593684683308807,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.2805,
"step": 760
},
{
"epoch": 3.0855696202531644,
"grad_norm": 0.24857781563486514,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.2759,
"step": 761
},
{
"epoch": 3.089620253164557,
"grad_norm": 0.28342424836176744,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.2821,
"step": 762
},
{
"epoch": 3.0936708860759494,
"grad_norm": 0.23372599184749718,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2777,
"step": 763
},
{
"epoch": 3.097721518987342,
"grad_norm": 0.26044848663077724,
"learning_rate": 1.508449870007656e-05,
"loss": 0.2735,
"step": 764
},
{
"epoch": 3.101772151898734,
"grad_norm": 0.26954204091729994,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.2753,
"step": 765
},
{
"epoch": 3.1058227848101265,
"grad_norm": 0.2628228590424006,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.2782,
"step": 766
},
{
"epoch": 3.109873417721519,
"grad_norm": 0.28081125856806727,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2738,
"step": 767
},
{
"epoch": 3.1139240506329116,
"grad_norm": 0.26418922522374416,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.2772,
"step": 768
},
{
"epoch": 3.1179746835443036,
"grad_norm": 0.24828408761300716,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.276,
"step": 769
},
{
"epoch": 3.122025316455696,
"grad_norm": 0.26843828676847636,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2799,
"step": 770
},
{
"epoch": 3.1260759493670887,
"grad_norm": 0.22337059120647948,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.2676,
"step": 771
},
{
"epoch": 3.130126582278481,
"grad_norm": 0.2551643160802219,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.2708,
"step": 772
},
{
"epoch": 3.1341772151898732,
"grad_norm": 0.2374420200265584,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.2815,
"step": 773
},
{
"epoch": 3.1382278481012658,
"grad_norm": 0.24253983107767463,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2811,
"step": 774
},
{
"epoch": 3.1422784810126583,
"grad_norm": 0.21930775094832408,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.2795,
"step": 775
},
{
"epoch": 3.146329113924051,
"grad_norm": 0.25879877941070084,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2786,
"step": 776
},
{
"epoch": 3.150379746835443,
"grad_norm": 0.22014425393164447,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.274,
"step": 777
},
{
"epoch": 3.1544303797468354,
"grad_norm": 0.25864198804981503,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.2759,
"step": 778
},
{
"epoch": 3.158481012658228,
"grad_norm": 0.2193494655023655,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2674,
"step": 779
},
{
"epoch": 3.1625316455696204,
"grad_norm": 0.23157070125424858,
"learning_rate": 1.420958430225303e-05,
"loss": 0.2727,
"step": 780
},
{
"epoch": 3.1665822784810125,
"grad_norm": 0.249535937176069,
"learning_rate": 1.415527989393985e-05,
"loss": 0.2791,
"step": 781
},
{
"epoch": 3.170632911392405,
"grad_norm": 0.24725453646952017,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2768,
"step": 782
},
{
"epoch": 3.1746835443037975,
"grad_norm": 0.2649662856698457,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2833,
"step": 783
},
{
"epoch": 3.17873417721519,
"grad_norm": 0.25154628213813546,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2774,
"step": 784
},
{
"epoch": 3.182784810126582,
"grad_norm": 0.23734565400914417,
"learning_rate": 1.393853735430398e-05,
"loss": 0.2805,
"step": 785
},
{
"epoch": 3.1868354430379746,
"grad_norm": 0.25943683789106914,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.2741,
"step": 786
},
{
"epoch": 3.190886075949367,
"grad_norm": 0.22496539720027176,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2841,
"step": 787
},
{
"epoch": 3.1949367088607596,
"grad_norm": 0.2243672364673809,
"learning_rate": 1.377649151173163e-05,
"loss": 0.2714,
"step": 788
},
{
"epoch": 3.1989873417721517,
"grad_norm": 0.2345541629468349,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.275,
"step": 789
},
{
"epoch": 3.2030379746835442,
"grad_norm": 0.2404565231528039,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.271,
"step": 790
},
{
"epoch": 3.2070886075949367,
"grad_norm": 0.2343771924736244,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2773,
"step": 791
},
{
"epoch": 3.2111392405063293,
"grad_norm": 0.23456504479677648,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.2849,
"step": 792
},
{
"epoch": 3.2151898734177213,
"grad_norm": 0.25731148265134773,
"learning_rate": 1.350742335331241e-05,
"loss": 0.2748,
"step": 793
},
{
"epoch": 3.219240506329114,
"grad_norm": 0.23295344245025082,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2844,
"step": 794
},
{
"epoch": 3.2232911392405064,
"grad_norm": 0.23663562664408433,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.275,
"step": 795
},
{
"epoch": 3.227341772151899,
"grad_norm": 0.2666206824062976,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.279,
"step": 796
},
{
"epoch": 3.231392405063291,
"grad_norm": 0.2161342848699084,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.2701,
"step": 797
},
{
"epoch": 3.2354430379746835,
"grad_norm": 0.2514435616549507,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2794,
"step": 798
},
{
"epoch": 3.239493670886076,
"grad_norm": 0.25730395942139267,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2795,
"step": 799
},
{
"epoch": 3.2435443037974685,
"grad_norm": 0.2498543315906576,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.2645,
"step": 800
},
{
"epoch": 3.2475949367088606,
"grad_norm": 0.22882769350849627,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2706,
"step": 801
},
{
"epoch": 3.251645569620253,
"grad_norm": 0.24810833524760129,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2722,
"step": 802
},
{
"epoch": 3.2556962025316456,
"grad_norm": 0.21842894967216595,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.2793,
"step": 803
},
{
"epoch": 3.259746835443038,
"grad_norm": 0.2373757028402876,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.2788,
"step": 804
},
{
"epoch": 3.26379746835443,
"grad_norm": 0.2405878006158352,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.2842,
"step": 805
},
{
"epoch": 3.2678481012658227,
"grad_norm": 0.2433827387599161,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2625,
"step": 806
},
{
"epoch": 3.271898734177215,
"grad_norm": 0.24913909388591932,
"learning_rate": 1.276115892225764e-05,
"loss": 0.2751,
"step": 807
},
{
"epoch": 3.2759493670886077,
"grad_norm": 0.23626673184553668,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.2748,
"step": 808
},
{
"epoch": 3.2800000000000002,
"grad_norm": 0.24423231022856146,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.2754,
"step": 809
},
{
"epoch": 3.2840506329113923,
"grad_norm": 0.2561797393500844,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2787,
"step": 810
},
{
"epoch": 3.288101265822785,
"grad_norm": 0.2207431577189347,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.2823,
"step": 811
},
{
"epoch": 3.2921518987341774,
"grad_norm": 0.2390607698975217,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2772,
"step": 812
},
{
"epoch": 3.2962025316455694,
"grad_norm": 0.22646536385285418,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.2682,
"step": 813
},
{
"epoch": 3.300253164556962,
"grad_norm": 0.23286613465553463,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2742,
"step": 814
},
{
"epoch": 3.3043037974683545,
"grad_norm": 0.21906991767999817,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.2693,
"step": 815
},
{
"epoch": 3.308354430379747,
"grad_norm": 0.2052730434574235,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.2716,
"step": 816
},
{
"epoch": 3.3124050632911395,
"grad_norm": 0.22400366863194632,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2745,
"step": 817
},
{
"epoch": 3.3164556962025316,
"grad_norm": 0.22708700242804397,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2786,
"step": 818
},
{
"epoch": 3.320506329113924,
"grad_norm": 0.22857308465727585,
"learning_rate": 1.213055323844297e-05,
"loss": 0.2702,
"step": 819
},
{
"epoch": 3.3245569620253166,
"grad_norm": 0.22550351828643592,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2732,
"step": 820
},
{
"epoch": 3.3286075949367087,
"grad_norm": 0.21763456474769924,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2801,
"step": 821
},
{
"epoch": 3.332658227848101,
"grad_norm": 0.22143383847100911,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.2792,
"step": 822
},
{
"epoch": 3.3367088607594937,
"grad_norm": 0.2238284401396544,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2856,
"step": 823
},
{
"epoch": 3.340759493670886,
"grad_norm": 0.2120369517204791,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2811,
"step": 824
},
{
"epoch": 3.3448101265822787,
"grad_norm": 0.2141708114077099,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.2847,
"step": 825
},
{
"epoch": 3.348860759493671,
"grad_norm": 0.20380182735456218,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.2748,
"step": 826
},
{
"epoch": 3.3529113924050633,
"grad_norm": 0.21537578360155757,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2831,
"step": 827
},
{
"epoch": 3.356962025316456,
"grad_norm": 0.21684343955338886,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.2641,
"step": 828
},
{
"epoch": 3.361012658227848,
"grad_norm": 0.2053014326977809,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2892,
"step": 829
},
{
"epoch": 3.3650632911392404,
"grad_norm": 0.22461574504255327,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2737,
"step": 830
},
{
"epoch": 3.369113924050633,
"grad_norm": 0.2059273927675095,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.2759,
"step": 831
},
{
"epoch": 3.3731645569620254,
"grad_norm": 0.21025435160238648,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2717,
"step": 832
},
{
"epoch": 3.377215189873418,
"grad_norm": 0.2026407430207601,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2799,
"step": 833
},
{
"epoch": 3.38126582278481,
"grad_norm": 0.21059791350955348,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2837,
"step": 834
},
{
"epoch": 3.3853164556962025,
"grad_norm": 0.22416042332216776,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.2787,
"step": 835
},
{
"epoch": 3.389367088607595,
"grad_norm": 0.22554264474832128,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2763,
"step": 836
},
{
"epoch": 3.393417721518987,
"grad_norm": 0.2199436585029525,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2775,
"step": 837
},
{
"epoch": 3.3974683544303796,
"grad_norm": 0.2075334489470918,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.2691,
"step": 838
},
{
"epoch": 3.401518987341772,
"grad_norm": 0.22531217217688382,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.2744,
"step": 839
},
{
"epoch": 3.4055696202531647,
"grad_norm": 0.20525061956030374,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2757,
"step": 840
},
{
"epoch": 3.409620253164557,
"grad_norm": 0.22242350929680185,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2805,
"step": 841
},
{
"epoch": 3.4136708860759493,
"grad_norm": 0.22948740601507822,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2841,
"step": 842
},
{
"epoch": 3.4177215189873418,
"grad_norm": 0.2042296237171055,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.2725,
"step": 843
},
{
"epoch": 3.4217721518987343,
"grad_norm": 0.2196917664512823,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.2709,
"step": 844
},
{
"epoch": 3.4258227848101264,
"grad_norm": 0.2216719844678615,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.2685,
"step": 845
},
{
"epoch": 3.429873417721519,
"grad_norm": 0.20862589258910247,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2733,
"step": 846
},
{
"epoch": 3.4339240506329114,
"grad_norm": 0.1936718821836611,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.2675,
"step": 847
},
{
"epoch": 3.437974683544304,
"grad_norm": 0.21227077727776492,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2733,
"step": 848
},
{
"epoch": 3.4420253164556964,
"grad_norm": 0.1971891607834445,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2714,
"step": 849
},
{
"epoch": 3.4460759493670885,
"grad_norm": 0.21094610545166867,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2788,
"step": 850
},
{
"epoch": 3.450126582278481,
"grad_norm": 0.2058258006833322,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.2818,
"step": 851
},
{
"epoch": 3.4541772151898735,
"grad_norm": 0.18601364232468848,
"learning_rate": 1.044560362975474e-05,
"loss": 0.2743,
"step": 852
},
{
"epoch": 3.4582278481012656,
"grad_norm": 0.21383681173618652,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2781,
"step": 853
},
{
"epoch": 3.462278481012658,
"grad_norm": 0.19856899049179957,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.278,
"step": 854
},
{
"epoch": 3.4663291139240506,
"grad_norm": 0.2074340611406711,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.2842,
"step": 855
},
{
"epoch": 3.470379746835443,
"grad_norm": 0.21994354737785393,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2757,
"step": 856
},
{
"epoch": 3.4744303797468357,
"grad_norm": 0.19626814259941,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.2762,
"step": 857
},
{
"epoch": 3.4784810126582277,
"grad_norm": 0.22254786637787546,
"learning_rate": 1.014782400334433e-05,
"loss": 0.2755,
"step": 858
},
{
"epoch": 3.4825316455696202,
"grad_norm": 0.22790430818939278,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2733,
"step": 859
},
{
"epoch": 3.4865822784810128,
"grad_norm": 0.22091438868120927,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2693,
"step": 860
},
{
"epoch": 3.490632911392405,
"grad_norm": 0.21488277437896963,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2753,
"step": 861
},
{
"epoch": 3.4946835443037974,
"grad_norm": 0.22218292110095883,
"learning_rate": 9.950885880977891e-06,
"loss": 0.2659,
"step": 862
},
{
"epoch": 3.49873417721519,
"grad_norm": 0.23012923666583673,
"learning_rate": 9.901852696139382e-06,
"loss": 0.2771,
"step": 863
},
{
"epoch": 3.5027848101265824,
"grad_norm": 0.20177014335263904,
"learning_rate": 9.852900840391027e-06,
"loss": 0.272,
"step": 864
},
{
"epoch": 3.506835443037975,
"grad_norm": 0.2290013620933438,
"learning_rate": 9.804030707984313e-06,
"loss": 0.2682,
"step": 865
},
{
"epoch": 3.510886075949367,
"grad_norm": 0.22643413023749495,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2774,
"step": 866
},
{
"epoch": 3.5149367088607595,
"grad_norm": 0.23019261796175292,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2716,
"step": 867
},
{
"epoch": 3.518987341772152,
"grad_norm": 0.21062586259078825,
"learning_rate": 9.657914583437454e-06,
"loss": 0.2736,
"step": 868
},
{
"epoch": 3.523037974683544,
"grad_norm": 0.2162884351315457,
"learning_rate": 9.609375273701246e-06,
"loss": 0.2781,
"step": 869
},
{
"epoch": 3.5270886075949366,
"grad_norm": 0.1970560773617703,
"learning_rate": 9.560919648628133e-06,
"loss": 0.2725,
"step": 870
},
{
"epoch": 3.531139240506329,
"grad_norm": 0.20296027289004667,
"learning_rate": 9.512548098473047e-06,
"loss": 0.2825,
"step": 871
},
{
"epoch": 3.5351898734177216,
"grad_norm": 0.19778681889061267,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2884,
"step": 872
},
{
"epoch": 3.539240506329114,
"grad_norm": 0.19710055539703294,
"learning_rate": 9.416058780547987e-06,
"loss": 0.2704,
"step": 873
},
{
"epoch": 3.543291139240506,
"grad_norm": 0.20499983031082322,
"learning_rate": 9.367941789889714e-06,
"loss": 0.2761,
"step": 874
},
{
"epoch": 3.5473417721518987,
"grad_norm": 0.19495119740439207,
"learning_rate": 9.319910428366607e-06,
"loss": 0.2826,
"step": 875
},
{
"epoch": 3.5513924050632912,
"grad_norm": 0.18794524503631538,
"learning_rate": 9.271965082816667e-06,
"loss": 0.2751,
"step": 876
},
{
"epoch": 3.5554430379746833,
"grad_norm": 0.20129018771063525,
"learning_rate": 9.224106139385111e-06,
"loss": 0.2651,
"step": 877
},
{
"epoch": 3.559493670886076,
"grad_norm": 0.19952501779511672,
"learning_rate": 9.176333983521291e-06,
"loss": 0.275,
"step": 878
},
{
"epoch": 3.5635443037974683,
"grad_norm": 0.20453381482082772,
"learning_rate": 9.12864899997558e-06,
"loss": 0.2739,
"step": 879
},
{
"epoch": 3.567594936708861,
"grad_norm": 0.20101993429331463,
"learning_rate": 9.08105157279628e-06,
"loss": 0.2768,
"step": 880
},
{
"epoch": 3.5716455696202534,
"grad_norm": 0.21602780497155846,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2772,
"step": 881
},
{
"epoch": 3.5756962025316454,
"grad_norm": 0.1833663657375999,
"learning_rate": 8.986120920201205e-06,
"loss": 0.2819,
"step": 882
},
{
"epoch": 3.579746835443038,
"grad_norm": 0.20185944702253625,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2733,
"step": 883
},
{
"epoch": 3.5837974683544305,
"grad_norm": 0.21067509602840082,
"learning_rate": 8.8915450839636e-06,
"loss": 0.285,
"step": 884
},
{
"epoch": 3.5878481012658225,
"grad_norm": 0.18808131322213165,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2738,
"step": 885
},
{
"epoch": 3.591898734177215,
"grad_norm": 0.20823634807470398,
"learning_rate": 8.797327110880479e-06,
"loss": 0.2784,
"step": 886
},
{
"epoch": 3.5959493670886076,
"grad_norm": 0.19497204382376102,
"learning_rate": 8.750353271996206e-06,
"loss": 0.2784,
"step": 887
},
{
"epoch": 3.6,
"grad_norm": 0.2125758947837265,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2812,
"step": 888
},
{
"epoch": 3.6040506329113926,
"grad_norm": 0.2223968699951753,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2789,
"step": 889
},
{
"epoch": 3.6081012658227847,
"grad_norm": 0.19356944156851066,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2825,
"step": 890
},
{
"epoch": 3.612151898734177,
"grad_norm": 0.19740250095773398,
"learning_rate": 8.563367719785698e-06,
"loss": 0.2825,
"step": 891
},
{
"epoch": 3.6162025316455697,
"grad_norm": 0.20654271940944888,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2806,
"step": 892
},
{
"epoch": 3.620253164556962,
"grad_norm": 0.19301394121491217,
"learning_rate": 8.47042609394269e-06,
"loss": 0.2681,
"step": 893
},
{
"epoch": 3.6243037974683543,
"grad_norm": 0.2080549465339316,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2754,
"step": 894
},
{
"epoch": 3.628354430379747,
"grad_norm": 0.23259066185797747,
"learning_rate": 8.37785589776465e-06,
"loss": 0.2845,
"step": 895
},
{
"epoch": 3.6324050632911393,
"grad_norm": 0.21081029207539165,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2773,
"step": 896
},
{
"epoch": 3.636455696202532,
"grad_norm": 0.20921167602487997,
"learning_rate": 8.285660113436104e-06,
"loss": 0.2757,
"step": 897
},
{
"epoch": 3.640506329113924,
"grad_norm": 0.2196673707250673,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2832,
"step": 898
},
{
"epoch": 3.6445569620253164,
"grad_norm": 0.18867302513068623,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2777,
"step": 899
},
{
"epoch": 3.648607594936709,
"grad_norm": 0.20453116259271406,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2753,
"step": 900
},
{
"epoch": 3.652658227848101,
"grad_norm": 0.21362123742741088,
"learning_rate": 8.102403648660859e-06,
"loss": 0.2779,
"step": 901
},
{
"epoch": 3.6567088607594935,
"grad_norm": 0.18961913585527132,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2757,
"step": 902
},
{
"epoch": 3.660759493670886,
"grad_norm": 0.19972501626354222,
"learning_rate": 8.011348871891762e-06,
"loss": 0.2766,
"step": 903
},
{
"epoch": 3.6648101265822786,
"grad_norm": 0.20847154609254703,
"learning_rate": 7.965966132787287e-06,
"loss": 0.2735,
"step": 904
},
{
"epoch": 3.668860759493671,
"grad_norm": 0.20105463408852836,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2799,
"step": 905
},
{
"epoch": 3.672911392405063,
"grad_norm": 0.20103777360526406,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2794,
"step": 906
},
{
"epoch": 3.6769620253164557,
"grad_norm": 0.19480517741053335,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2685,
"step": 907
},
{
"epoch": 3.681012658227848,
"grad_norm": 0.21161874797494806,
"learning_rate": 7.785408024250259e-06,
"loss": 0.2745,
"step": 908
},
{
"epoch": 3.6850632911392402,
"grad_norm": 0.1995021810978668,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2728,
"step": 909
},
{
"epoch": 3.6891139240506328,
"grad_norm": 0.1888519208740877,
"learning_rate": 7.695717765619257e-06,
"loss": 0.2728,
"step": 910
},
{
"epoch": 3.6931645569620253,
"grad_norm": 0.21218285823909927,
"learning_rate": 7.651021101410673e-06,
"loss": 0.2752,
"step": 911
},
{
"epoch": 3.697215189873418,
"grad_norm": 0.20291390610069268,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2703,
"step": 912
},
{
"epoch": 3.7012658227848103,
"grad_norm": 0.18496461753553606,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2834,
"step": 913
},
{
"epoch": 3.7053164556962024,
"grad_norm": 0.20101868879086876,
"learning_rate": 7.517529286570349e-06,
"loss": 0.2773,
"step": 914
},
{
"epoch": 3.709367088607595,
"grad_norm": 0.18936592781896391,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2754,
"step": 915
},
{
"epoch": 3.7134177215189874,
"grad_norm": 0.18813999159437855,
"learning_rate": 7.429036806562935e-06,
"loss": 0.2681,
"step": 916
},
{
"epoch": 3.7174683544303795,
"grad_norm": 0.19953784770470734,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2685,
"step": 917
},
{
"epoch": 3.721518987341772,
"grad_norm": 0.1796340930559221,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2775,
"step": 918
},
{
"epoch": 3.7255696202531645,
"grad_norm": 0.1996742276240275,
"learning_rate": 7.297058308409282e-06,
"loss": 0.2777,
"step": 919
},
{
"epoch": 3.729620253164557,
"grad_norm": 0.1978506596272937,
"learning_rate": 7.25326961958704e-06,
"loss": 0.2697,
"step": 920
},
{
"epoch": 3.7336708860759495,
"grad_norm": 0.1867924695303605,
"learning_rate": 7.209583591178921e-06,
"loss": 0.2842,
"step": 921
},
{
"epoch": 3.7377215189873416,
"grad_norm": 0.2035882953953522,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2735,
"step": 922
},
{
"epoch": 3.741772151898734,
"grad_norm": 0.2009378197369855,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.28,
"step": 923
},
{
"epoch": 3.7458227848101266,
"grad_norm": 0.1859553811994067,
"learning_rate": 7.079144982700909e-06,
"loss": 0.2749,
"step": 924
},
{
"epoch": 3.7498734177215187,
"grad_norm": 0.1877575853453917,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.2816,
"step": 925
},
{
"epoch": 3.7539240506329112,
"grad_norm": 0.19286013807772873,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2779,
"step": 926
},
{
"epoch": 3.7579746835443038,
"grad_norm": 0.18093429282752174,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2609,
"step": 927
},
{
"epoch": 3.7620253164556963,
"grad_norm": 0.2024501836810843,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2734,
"step": 928
},
{
"epoch": 3.766075949367089,
"grad_norm": 0.1935109436260187,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2752,
"step": 929
},
{
"epoch": 3.770126582278481,
"grad_norm": 0.20059496519236814,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2787,
"step": 930
},
{
"epoch": 3.7741772151898734,
"grad_norm": 0.19466258537767767,
"learning_rate": 6.77844662204546e-06,
"loss": 0.2753,
"step": 931
},
{
"epoch": 3.778227848101266,
"grad_norm": 0.20987834453082269,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2758,
"step": 932
},
{
"epoch": 3.782278481012658,
"grad_norm": 0.20453234989759053,
"learning_rate": 6.693485996688695e-06,
"loss": 0.2697,
"step": 933
},
{
"epoch": 3.7863291139240505,
"grad_norm": 0.19490058039534425,
"learning_rate": 6.651166266411801e-06,
"loss": 0.2815,
"step": 934
},
{
"epoch": 3.790379746835443,
"grad_norm": 0.20778284574113687,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2765,
"step": 935
},
{
"epoch": 3.7944303797468355,
"grad_norm": 0.19461485892360433,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2767,
"step": 936
},
{
"epoch": 3.798481012658228,
"grad_norm": 0.18805828828117163,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2743,
"step": 937
},
{
"epoch": 3.80253164556962,
"grad_norm": 0.1961008349642996,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.2717,
"step": 938
},
{
"epoch": 3.8065822784810126,
"grad_norm": 0.1837406696274524,
"learning_rate": 6.441187046486648e-06,
"loss": 0.2697,
"step": 939
},
{
"epoch": 3.810632911392405,
"grad_norm": 0.18668064598493148,
"learning_rate": 6.399517456398567e-06,
"loss": 0.268,
"step": 940
},
{
"epoch": 3.814683544303797,
"grad_norm": 0.20026993691714626,
"learning_rate": 6.357957402727164e-06,
"loss": 0.2777,
"step": 941
},
{
"epoch": 3.81873417721519,
"grad_norm": 0.18999937800237127,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2877,
"step": 942
},
{
"epoch": 3.8227848101265822,
"grad_norm": 0.18902044048917935,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2647,
"step": 943
},
{
"epoch": 3.8268354430379747,
"grad_norm": 0.19778415331528262,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2767,
"step": 944
},
{
"epoch": 3.8308860759493673,
"grad_norm": 0.18815771844504509,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2768,
"step": 945
},
{
"epoch": 3.8349367088607593,
"grad_norm": 0.19555391500386574,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2721,
"step": 946
},
{
"epoch": 3.838987341772152,
"grad_norm": 0.20187932675565506,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.266,
"step": 947
},
{
"epoch": 3.8430379746835444,
"grad_norm": 0.18680949565862998,
"learning_rate": 6.070132051252868e-06,
"loss": 0.2687,
"step": 948
},
{
"epoch": 3.847088607594937,
"grad_norm": 0.19083347464479936,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2792,
"step": 949
},
{
"epoch": 3.8511392405063294,
"grad_norm": 0.19930222271192738,
"learning_rate": 5.988900993811575e-06,
"loss": 0.2784,
"step": 950
},
{
"epoch": 3.8551898734177215,
"grad_norm": 0.19309186243086163,
"learning_rate": 5.948454566959363e-06,
"loss": 0.277,
"step": 951
},
{
"epoch": 3.859240506329114,
"grad_norm": 0.19703190225927963,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2821,
"step": 952
},
{
"epoch": 3.8632911392405065,
"grad_norm": 0.19593183032332265,
"learning_rate": 5.867901545738976e-06,
"loss": 0.2836,
"step": 953
},
{
"epoch": 3.8673417721518986,
"grad_norm": 0.21213989881849907,
"learning_rate": 5.827795600133774e-06,
"loss": 0.2768,
"step": 954
},
{
"epoch": 3.871392405063291,
"grad_norm": 0.2002854974246593,
"learning_rate": 5.787803795514466e-06,
"loss": 0.2776,
"step": 955
},
{
"epoch": 3.8754430379746836,
"grad_norm": 0.1911363382535373,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2752,
"step": 956
},
{
"epoch": 3.879493670886076,
"grad_norm": 0.20053490755737824,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2643,
"step": 957
},
{
"epoch": 3.8835443037974686,
"grad_norm": 0.1834167606626894,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2703,
"step": 958
},
{
"epoch": 3.8875949367088607,
"grad_norm": 0.18672195437661768,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2788,
"step": 959
},
{
"epoch": 3.891645569620253,
"grad_norm": 0.1901818443137941,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.2805,
"step": 960
},
{
"epoch": 3.8956962025316457,
"grad_norm": 0.19548926088334329,
"learning_rate": 5.550267900783019e-06,
"loss": 0.2782,
"step": 961
},
{
"epoch": 3.899746835443038,
"grad_norm": 0.19966278704246876,
"learning_rate": 5.511084049737623e-06,
"loss": 0.2856,
"step": 962
},
{
"epoch": 3.9037974683544303,
"grad_norm": 0.2008166996365627,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2698,
"step": 963
},
{
"epoch": 3.907848101265823,
"grad_norm": 0.21148191767085694,
"learning_rate": 5.433066737499948e-06,
"loss": 0.2752,
"step": 964
},
{
"epoch": 3.9118987341772153,
"grad_norm": 0.19068784613258674,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2777,
"step": 965
},
{
"epoch": 3.915949367088608,
"grad_norm": 0.19945851084337807,
"learning_rate": 5.355518704628997e-06,
"loss": 0.2743,
"step": 966
},
{
"epoch": 3.92,
"grad_norm": 0.2136641421303689,
"learning_rate": 5.316921449248731e-06,
"loss": 0.2755,
"step": 967
},
{
"epoch": 3.9240506329113924,
"grad_norm": 0.203941473642444,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2739,
"step": 968
},
{
"epoch": 3.928101265822785,
"grad_norm": 0.18132375132160297,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2791,
"step": 969
},
{
"epoch": 3.932151898734177,
"grad_norm": 0.19852684015531197,
"learning_rate": 5.201840454747822e-06,
"loss": 0.2714,
"step": 970
},
{
"epoch": 3.9362025316455695,
"grad_norm": 0.21443985385347494,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2789,
"step": 971
},
{
"epoch": 3.940253164556962,
"grad_norm": 0.1893615030108217,
"learning_rate": 5.125715188541609e-06,
"loss": 0.2781,
"step": 972
},
{
"epoch": 3.9443037974683546,
"grad_norm": 0.18910347472282493,
"learning_rate": 5.087832095571021e-06,
"loss": 0.2833,
"step": 973
},
{
"epoch": 3.948354430379747,
"grad_norm": 0.18729041268791086,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.279,
"step": 974
},
{
"epoch": 3.952405063291139,
"grad_norm": 0.20535495755859548,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2858,
"step": 975
},
{
"epoch": 3.9564556962025317,
"grad_norm": 0.1916126806667645,
"learning_rate": 4.974904635542815e-06,
"loss": 0.274,
"step": 976
},
{
"epoch": 3.960506329113924,
"grad_norm": 0.1794470948096593,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2809,
"step": 977
},
{
"epoch": 3.9645569620253163,
"grad_norm": 0.19377313245843622,
"learning_rate": 4.900224207170299e-06,
"loss": 0.2799,
"step": 978
},
{
"epoch": 3.968607594936709,
"grad_norm": 0.20693629297330265,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2753,
"step": 979
},
{
"epoch": 3.9726582278481013,
"grad_norm": 0.177707972610654,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.2781,
"step": 980
},
{
"epoch": 3.976708860759494,
"grad_norm": 0.18281793637213134,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2662,
"step": 981
},
{
"epoch": 3.9807594936708863,
"grad_norm": 0.19549982826554674,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2787,
"step": 982
},
{
"epoch": 3.9848101265822784,
"grad_norm": 0.19757475568032545,
"learning_rate": 4.715656557757473e-06,
"loss": 0.2771,
"step": 983
},
{
"epoch": 3.988860759493671,
"grad_norm": 0.19000737460753786,
"learning_rate": 4.679111137620442e-06,
"loss": 0.2712,
"step": 984
},
{
"epoch": 3.9929113924050634,
"grad_norm": 0.19040360577540066,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2783,
"step": 985
},
{
"epoch": 3.9969620253164555,
"grad_norm": 0.20008438292731473,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2723,
"step": 986
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.2748580290853044,
"learning_rate": 4.570216403561141e-06,
"loss": 0.2607,
"step": 987
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.3181883350953151,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2478,
"step": 988
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.2178856577601294,
"learning_rate": 4.498240773728859e-06,
"loss": 0.2326,
"step": 989
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.36187954226473223,
"learning_rate": 4.462440087699609e-06,
"loss": 0.2297,
"step": 990
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.38231276398938474,
"learning_rate": 4.426764539041861e-06,
"loss": 0.2345,
"step": 991
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.22313708491244713,
"learning_rate": 4.391214415081582e-06,
"loss": 0.2359,
"step": 992
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.32307829866786825,
"learning_rate": 4.355790002134579e-06,
"loss": 0.2383,
"step": 993
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.31741867046814415,
"learning_rate": 4.320491585504207e-06,
"loss": 0.2374,
"step": 994
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.2700877927015394,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.2333,
"step": 995
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.22969558204435947,
"learning_rate": 4.250273877330691e-06,
"loss": 0.2472,
"step": 996
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.2631832606536734,
"learning_rate": 4.215355151311313e-06,
"loss": 0.2467,
"step": 997
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.24548744896455704,
"learning_rate": 4.180563552651542e-06,
"loss": 0.2347,
"step": 998
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.22511668596012968,
"learning_rate": 4.145899361558147e-06,
"loss": 0.2449,
"step": 999
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.2135192091005289,
"learning_rate": 4.111362857211738e-06,
"loss": 0.2401,
"step": 1000
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.23284287715776902,
"learning_rate": 4.076954317764592e-06,
"loss": 0.241,
"step": 1001
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.22490275820256497,
"learning_rate": 4.042674020338335e-06,
"loss": 0.2378,
"step": 1002
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.20667559516771677,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.239,
"step": 1003
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.21234646235388502,
"learning_rate": 3.974499254868674e-06,
"loss": 0.244,
"step": 1004
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.21644038526346188,
"learning_rate": 3.940605335895451e-06,
"loss": 0.231,
"step": 1005
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.2099004635120715,
"learning_rate": 3.90684075707908e-06,
"loss": 0.2346,
"step": 1006
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.2022363835992099,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.2403,
"step": 1007
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.19694675452546928,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.2452,
"step": 1008
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.19899492564637836,
"learning_rate": 3.806325775702304e-06,
"loss": 0.2477,
"step": 1009
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.19936067428069223,
"learning_rate": 3.773081266416434e-06,
"loss": 0.2387,
"step": 1010
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.19303684863177623,
"learning_rate": 3.739967446503245e-06,
"loss": 0.2454,
"step": 1011
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.2025412313727815,
"learning_rate": 3.706984582656894e-06,
"loss": 0.2536,
"step": 1012
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.19199666507505275,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.2349,
"step": 1013
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.197263111520628,
"learning_rate": 3.641412784665648e-06,
"loss": 0.2422,
"step": 1014
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.20369420883650075,
"learning_rate": 3.608824378627005e-06,
"loss": 0.237,
"step": 1015
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.18733524831047246,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.257,
"step": 1016
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.1841681429067796,
"learning_rate": 3.544043864774269e-06,
"loss": 0.239,
"step": 1017
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.1934412495884837,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.2302,
"step": 1018
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.20106064683336444,
"learning_rate": 3.479793485887819e-06,
"loss": 0.231,
"step": 1019
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.18829248656249797,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.2337,
"step": 1020
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.17806511485244234,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.238,
"step": 1021
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.18986146299153683,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.236,
"step": 1022
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.18867966722837776,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.2341,
"step": 1023
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.18683052693866892,
"learning_rate": 3.321500420335e-06,
"loss": 0.2339,
"step": 1024
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.1751703402207504,
"learning_rate": 3.290243771741275e-06,
"loss": 0.2322,
"step": 1025
},
{
"epoch": 4.16,
"grad_norm": 0.18223122207695241,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.2436,
"step": 1026
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.18929303651985566,
"learning_rate": 3.228134459445149e-06,
"loss": 0.2357,
"step": 1027
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.19553984554247925,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.2481,
"step": 1028
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.18395965995576466,
"learning_rate": 3.166565459258513e-06,
"loss": 0.2449,
"step": 1029
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.18498221523844388,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.2334,
"step": 1030
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.19598473529046787,
"learning_rate": 3.105538754650419e-06,
"loss": 0.2415,
"step": 1031
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.19269938053950353,
"learning_rate": 3.07522937824712e-06,
"loss": 0.2329,
"step": 1032
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.18436291716357112,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.2371,
"step": 1033
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.18091639347013436,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.233,
"step": 1034
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.18148200062074574,
"learning_rate": 2.985120078631465e-06,
"loss": 0.2372,
"step": 1035
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.18802482653420075,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.226,
"step": 1036
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.17375946360258546,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.2366,
"step": 1037
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.17672369475715133,
"learning_rate": 2.896244091927678e-06,
"loss": 0.24,
"step": 1038
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.18834363700840165,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.242,
"step": 1039
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.18292353857231272,
"learning_rate": 2.837681792951994e-06,
"loss": 0.2364,
"step": 1040
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.17855008932014832,
"learning_rate": 2.808607860255981e-06,
"loss": 0.2455,
"step": 1041
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.1801703831108559,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.2437,
"step": 1042
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.18207413708146372,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.2281,
"step": 1043
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.1785398726278458,
"learning_rate": 2.722217735873718e-06,
"loss": 0.2361,
"step": 1044
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.1799367763087164,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.2227,
"step": 1045
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.18032187753076975,
"learning_rate": 2.665319697490205e-06,
"loss": 0.2435,
"step": 1046
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.1804911646701487,
"learning_rate": 2.637079980714945e-06,
"loss": 0.2402,
"step": 1047
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.17716611265892956,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.2346,
"step": 1048
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.1774306039636593,
"learning_rate": 2.581020289189571e-06,
"loss": 0.2413,
"step": 1049
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.1798627139361783,
"learning_rate": 2.553200765936501e-06,
"loss": 0.2461,
"step": 1050
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.1739622014905373,
"learning_rate": 2.525521756805962e-06,
"loss": 0.2523,
"step": 1051
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.18457253078959102,
"learning_rate": 2.497983484720885e-06,
"loss": 0.2297,
"step": 1052
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.17275153556543676,
"learning_rate": 2.470586171470728e-06,
"loss": 0.2427,
"step": 1053
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.18025535050337435,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.238,
"step": 1054
},
{
"epoch": 4.27746835443038,
"grad_norm": 1.3370113206108944,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.2378,
"step": 1055
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.1742200260148851,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.243,
"step": 1056
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.17673347675625573,
"learning_rate": 2.362410902836978e-06,
"loss": 0.2392,
"step": 1057
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.17938384596797252,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.2418,
"step": 1058
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.17911032377696107,
"learning_rate": 2.309174704446411e-06,
"loss": 0.2386,
"step": 1059
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.17420210780072293,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.2402,
"step": 1060
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.1800948002961462,
"learning_rate": 2.256508422807855e-06,
"loss": 0.2384,
"step": 1061
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.17035737900900524,
"learning_rate": 2.230389531696946e-06,
"loss": 0.2384,
"step": 1062
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.17325195043477495,
"learning_rate": 2.204413754585857e-06,
"loss": 0.2307,
"step": 1063
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.175149681724095,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.2468,
"step": 1064
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.17339753498891267,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.2359,
"step": 1065
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.1883063310143349,
"learning_rate": 2.127347193531757e-06,
"loss": 0.2418,
"step": 1066
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.17475937572822545,
"learning_rate": 2.101945952921942e-06,
"loss": 0.2508,
"step": 1067
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.1668947154280704,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.2313,
"step": 1068
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.17775388643080375,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.2345,
"step": 1069
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.16949619418815015,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.2425,
"step": 1070
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.17067733987683373,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.2426,
"step": 1071
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.17843588950241027,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.2457,
"step": 1072
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.17128270267423293,
"learning_rate": 1.95257270594031e-06,
"loss": 0.2342,
"step": 1073
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.17207694884969532,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.2449,
"step": 1074
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.17685757029559587,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.2362,
"step": 1075
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.1773478128170538,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.232,
"step": 1076
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.17183838324685996,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.2255,
"step": 1077
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.17504069218411628,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.2368,
"step": 1078
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.17502860445127996,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.2368,
"step": 1079
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.18113540703274805,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.2395,
"step": 1080
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.17641544638175444,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.2384,
"step": 1081
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.17952464741387786,
"learning_rate": 1.738336950434061e-06,
"loss": 0.244,
"step": 1082
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.1693351243866881,
"learning_rate": 1.715265841375957e-06,
"loss": 0.2399,
"step": 1083
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.17256222803863375,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.2314,
"step": 1084
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.17362962860270104,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.2398,
"step": 1085
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.1677505898744511,
"learning_rate": 1.646936827927441e-06,
"loss": 0.2355,
"step": 1086
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.17232762940044918,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.2336,
"step": 1087
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.16919108159395574,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.235,
"step": 1088
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.1747907407878458,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.2216,
"step": 1089
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.17045686468223906,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.2508,
"step": 1090
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.18124463012338662,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.2395,
"step": 1091
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.1789171074211448,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.2349,
"step": 1092
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.20750967563872047,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.2382,
"step": 1093
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.17050954313653252,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.2435,
"step": 1094
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.18780731101699646,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.2413,
"step": 1095
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.1744674574073054,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.2373,
"step": 1096
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.16589544124773883,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.2447,
"step": 1097
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.17731317148319564,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.2361,
"step": 1098
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.1790779519400896,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.2442,
"step": 1099
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.17814567615026636,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.2421,
"step": 1100
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.17796295881912252,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.2334,
"step": 1101
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.17311187148031132,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.2405,
"step": 1102
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.1782908605972588,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.2461,
"step": 1103
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.1777648129839316,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.2444,
"step": 1104
},
{
"epoch": 4.48,
"grad_norm": 0.17550733068562901,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.2397,
"step": 1105
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.17560412034741613,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.2445,
"step": 1106
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.17861752145468918,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.2286,
"step": 1107
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.16483977878539505,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.2305,
"step": 1108
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.17439529177375232,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.2382,
"step": 1109
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.17253436837986533,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.2447,
"step": 1110
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.17570209011249804,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.2426,
"step": 1111
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.17280000109667432,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.2373,
"step": 1112
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.172074238115642,
"learning_rate": 1.092402023876933e-06,
"loss": 0.2393,
"step": 1113
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.16835901361487496,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.238,
"step": 1114
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.170560495459622,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.2294,
"step": 1115
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.1770138060104142,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.2434,
"step": 1116
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.1790890507480007,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.246,
"step": 1117
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.16654641421344335,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.2288,
"step": 1118
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.1724748505219568,
"learning_rate": 9.841383857091947e-07,
"loss": 0.2375,
"step": 1119
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.17250197067435233,
"learning_rate": 9.666296380321616e-07,
"loss": 0.2425,
"step": 1120
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.16598339402452747,
"learning_rate": 9.492741825034124e-07,
"loss": 0.2284,
"step": 1121
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.16866119920521755,
"learning_rate": 9.320721589013892e-07,
"loss": 0.2401,
"step": 1122
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.1766269819623834,
"learning_rate": 9.150237057688339e-07,
"loss": 0.2457,
"step": 1123
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.18256032190154015,
"learning_rate": 8.981289604116328e-07,
"loss": 0.2325,
"step": 1124
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.16275703067681954,
"learning_rate": 8.813880588977542e-07,
"loss": 0.2331,
"step": 1125
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.17184769303563996,
"learning_rate": 8.648011360561126e-07,
"loss": 0.2362,
"step": 1126
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.17469401235210877,
"learning_rate": 8.483683254755037e-07,
"loss": 0.2348,
"step": 1127
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.18268971230599423,
"learning_rate": 8.320897595035227e-07,
"loss": 0.244,
"step": 1128
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.17221918885177834,
"learning_rate": 8.159655692455093e-07,
"loss": 0.2295,
"step": 1129
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.16695762643832016,
"learning_rate": 7.999958845634648e-07,
"loss": 0.243,
"step": 1130
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.1760499670989461,
"learning_rate": 7.841808340750478e-07,
"loss": 0.2332,
"step": 1131
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.16807271058103718,
"learning_rate": 7.685205451524869e-07,
"loss": 0.2408,
"step": 1132
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.17092788872782386,
"learning_rate": 7.530151439216027e-07,
"loss": 0.2419,
"step": 1133
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.17337393035926343,
"learning_rate": 7.376647552607675e-07,
"loss": 0.2315,
"step": 1134
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.17475042752030567,
"learning_rate": 7.224695027998963e-07,
"loss": 0.2322,
"step": 1135
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.1776230414565918,
"learning_rate": 7.07429508919466e-07,
"loss": 0.2367,
"step": 1136
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.17024643283355934,
"learning_rate": 6.925448947495206e-07,
"loss": 0.2348,
"step": 1137
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.17278456100398357,
"learning_rate": 6.778157801686936e-07,
"loss": 0.2321,
"step": 1138
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.17426285503262076,
"learning_rate": 6.632422838032515e-07,
"loss": 0.2453,
"step": 1139
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.16871397330817198,
"learning_rate": 6.488245230261281e-07,
"loss": 0.2329,
"step": 1140
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.17043610869397036,
"learning_rate": 6.345626139559868e-07,
"loss": 0.2458,
"step": 1141
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.16858705692641293,
"learning_rate": 6.204566714562866e-07,
"loss": 0.2408,
"step": 1142
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.16701785414789116,
"learning_rate": 6.06506809134344e-07,
"loss": 0.2478,
"step": 1143
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.16835518776549152,
"learning_rate": 5.927131393404373e-07,
"loss": 0.2462,
"step": 1144
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.16278115715924493,
"learning_rate": 5.790757731668817e-07,
"loss": 0.2405,
"step": 1145
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.17418340588768064,
"learning_rate": 5.655948204471507e-07,
"loss": 0.2395,
"step": 1146
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.17191744125752978,
"learning_rate": 5.522703897549875e-07,
"loss": 0.238,
"step": 1147
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.17041561615958434,
"learning_rate": 5.391025884035239e-07,
"loss": 0.2483,
"step": 1148
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.17133055357476076,
"learning_rate": 5.260915224444207e-07,
"loss": 0.2337,
"step": 1149
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.1639244699237958,
"learning_rate": 5.132372966670129e-07,
"loss": 0.2325,
"step": 1150
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.17027711366048773,
"learning_rate": 5.005400145974704e-07,
"loss": 0.2448,
"step": 1151
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.16876584830155184,
"learning_rate": 4.879997784979562e-07,
"loss": 0.247,
"step": 1152
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.16278402947076076,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.2452,
"step": 1153
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.1662459165366728,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.2391,
"step": 1154
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.17378714797672973,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.2491,
"step": 1155
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.165230415238493,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.2384,
"step": 1156
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.16821344864301868,
"learning_rate": 4.27657778143431e-07,
"loss": 0.2331,
"step": 1157
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.16229292978487156,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.2478,
"step": 1158
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.16661581549688181,
"learning_rate": 4.046237371675177e-07,
"loss": 0.2325,
"step": 1159
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.1714307697811041,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.2313,
"step": 1160
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.16849212281125148,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.2425,
"step": 1161
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.17559950955387607,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.246,
"step": 1162
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.17036795689641485,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.2456,
"step": 1163
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.16949212847518805,
"learning_rate": 3.498020255137813e-07,
"loss": 0.2335,
"step": 1164
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.16817617726752535,
"learning_rate": 3.393121163057811e-07,
"loss": 0.2407,
"step": 1165
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.1639207339443521,
"learning_rate": 3.289805515751399e-07,
"loss": 0.2388,
"step": 1166
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.16383610426613418,
"learning_rate": 3.188074145308573e-07,
"loss": 0.2483,
"step": 1167
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.16611421420199685,
"learning_rate": 3.087927871059804e-07,
"loss": 0.2348,
"step": 1168
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.1696026079160522,
"learning_rate": 2.989367499569418e-07,
"loss": 0.2268,
"step": 1169
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.16709641974825176,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.2391,
"step": 1170
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.16317680048019856,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.2439,
"step": 1171
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.1637235693621484,
"learning_rate": 2.703209675663887e-07,
"loss": 0.236,
"step": 1172
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.1671758318361993,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.234,
"step": 1173
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.1652941458648819,
"learning_rate": 2.520381518804471e-07,
"loss": 0.246,
"step": 1174
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.16535588194124762,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.2327,
"step": 1175
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.1673933910133411,
"learning_rate": 2.343915243930317e-07,
"loss": 0.2297,
"step": 1176
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.1657833929975564,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.2365,
"step": 1177
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.16975206405643561,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.2342,
"step": 1178
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.16622413286978577,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.2321,
"step": 1179
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.16466181833137464,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.2431,
"step": 1180
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.16946749266293676,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.2421,
"step": 1181
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.17100778486146992,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.2385,
"step": 1182
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.16895883748053409,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.2362,
"step": 1183
},
{
"epoch": 4.8,
"grad_norm": 0.17008601722711325,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.241,
"step": 1184
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.16055709173650992,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.2403,
"step": 1185
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.1743071376644176,
"learning_rate": 1.557203888516745e-07,
"loss": 0.2445,
"step": 1186
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.168028284694555,
"learning_rate": 1.487313098987131e-07,
"loss": 0.228,
"step": 1187
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.16853481351271485,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.2372,
"step": 1188
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.16785777142342054,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.2408,
"step": 1189
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.16860713845930417,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.2484,
"step": 1190
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.16538633863255942,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.2328,
"step": 1191
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.1688151454668315,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.2304,
"step": 1192
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.16220584638954547,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.2359,
"step": 1193
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.17016362927783346,
"learning_rate": 1.042873700093061e-07,
"loss": 0.2425,
"step": 1194
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.1709690304958774,
"learning_rate": 9.857879404556291e-08,
"loss": 0.2471,
"step": 1195
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.16637626652803017,
"learning_rate": 9.303050139225722e-08,
"loss": 0.2434,
"step": 1196
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.16036914875632738,
"learning_rate": 8.76425367345779e-08,
"loss": 0.2443,
"step": 1197
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.1808542928501649,
"learning_rate": 8.241494346644897e-08,
"loss": 0.2373,
"step": 1198
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.16320078841617214,
"learning_rate": 7.734776369019204e-08,
"loss": 0.2508,
"step": 1199
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.1726016093453622,
"learning_rate": 7.244103821617332e-08,
"loss": 0.2434,
"step": 1200
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.16417355831521144,
"learning_rate": 6.769480656248606e-08,
"loss": 0.2387,
"step": 1201
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.16884412008512656,
"learning_rate": 6.310910695462635e-08,
"loss": 0.2443,
"step": 1202
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.16613864812034165,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.2446,
"step": 1203
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.16340694891146293,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.2308,
"step": 1204
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.16393755618855363,
"learning_rate": 5.031556326567488e-08,
"loss": 0.2281,
"step": 1205
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.16955183032873136,
"learning_rate": 4.637234823364312e-08,
"loss": 0.2431,
"step": 1206
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.17124135639952123,
"learning_rate": 4.258983697558838e-08,
"loss": 0.2345,
"step": 1207
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.17085745518038264,
"learning_rate": 3.896805995533548e-08,
"loss": 0.2409,
"step": 1208
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.16894238366570616,
"learning_rate": 3.550704634218028e-08,
"loss": 0.2331,
"step": 1209
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.16119563320419789,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.2491,
"step": 1210
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.16707297209169517,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.2319,
"step": 1211
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.1642809725116477,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.2426,
"step": 1212
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.17082453393406213,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.2356,
"step": 1213
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.1655965710119191,
"learning_rate": 2.061435966221881e-08,
"loss": 0.2412,
"step": 1214
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.16711054826283725,
"learning_rate": 1.811846652477245e-08,
"loss": 0.2385,
"step": 1215
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.15666741791881,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.2389,
"step": 1216
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.16137906939480134,
"learning_rate": 1.360949303591097e-08,
"loss": 0.2414,
"step": 1217
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.1778768612328291,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.2303,
"step": 1218
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.16517722152511577,
"learning_rate": 9.744388816668172e-09,
"loss": 0.2424,
"step": 1219
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.16662359992575973,
"learning_rate": 8.05332740472009e-09,
"loss": 0.2276,
"step": 1220
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.16031209384180534,
"learning_rate": 6.523278382872811e-09,
"loss": 0.2419,
"step": 1221
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.16952367243425326,
"learning_rate": 5.15425407393133e-09,
"loss": 0.2369,
"step": 1222
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.16454682707201615,
"learning_rate": 3.94626550383137e-09,
"loss": 0.242,
"step": 1223
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.16648759209051425,
"learning_rate": 2.899322401546112e-09,
"loss": 0.2367,
"step": 1224
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.1661603537488957,
"learning_rate": 2.013433199010706e-09,
"loss": 0.2341,
"step": 1225
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.16541478768102652,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.2391,
"step": 1226
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.15712536607173913,
"learning_rate": 7.248437353468695e-10,
"loss": 0.2401,
"step": 1227
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.16474690987359106,
"learning_rate": 3.221538523412449e-10,
"loss": 0.2499,
"step": 1228
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.16555777743556863,
"learning_rate": 8.053862524670663e-11,
"loss": 0.2413,
"step": 1229
},
{
"epoch": 4.98632911392405,
"grad_norm": 0.16361699145151354,
"learning_rate": 0.0,
"loss": 0.2443,
"step": 1230
},
{
"epoch": 4.98632911392405,
"step": 1230,
"total_flos": 5.874068361932964e+18,
"train_loss": 0.3428490738679723,
"train_runtime": 54606.384,
"train_samples_per_second": 2.893,
"train_steps_per_second": 0.023
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.874068361932964e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}